# Difference between revisions of "PRE2017 3 Groep1"

## Project statement

### Concept

In the year 2032 the Mars One organization plans on landing the first humans on Mars and establish a permanent human colony there. The first humans will start to build an empire by building a base where people are able to live. After the first base is built, it will be developed further and further. In the end, people on Mars should be able to live there without any help from planet earth. Mars should become a self-sustainable environment.

Before we get there, a lot needs to happen. Right now, Mars does not have an atmosphere, so there is hardly any oxygen in the air. All of the water present on Mars is in the form of ice and lies far beneath the surface. Right now, no one is able to live there. Nothing is able to grow there and that needs to change.

In this project, we want to get a step closer to life on Mars. We want to do this by taking a look at the implementation of food production for the planet. We chose to focus on the plantation of potatoes, since potatoes are very nutrient and since there has already been done some research on potato plantation on Mars

For this purpose, we are going to create a fully autonomous greenhouse, which is able to plant and harvest food, to fertilize, and to water its crops. The greenhouse, which we are going to create, will consist of four walls and a rounded roof, all filled with a great amount of isolation material. The slightly tilted roof is necessary because of the fact that Mars has no atmosphere. Because there is no atmosphere, rocks and other junk from outer space will fall down on Mars, opposed to what happens on earth, where these rocks burn up in the atmosphere. With a slightly tilted roof, everything that falls on top of the roof will slide right off.

The isolation material in the walls and in the roof absorbs all the light coming in from outside the greenhouse. This way, an earthly cycle inside the greenhouse can be created. This will be done using LED lamps. The light will be controlled by a controller. An earthly cycle is favorable, because it is known that potatoes will grow under these circumstances.

Inside the greenhouse, the planting and harvesting will be done by an autonomous farming robot. This robot will be based on robots that are already being used on earth. The robot will look like a tractor. It will come in action only when it gets a signal that the potatoes need to be harvested. The robot will harvest the entire field and then plow the ground in the greenhouse. After the ground has been fertilized again, the robot will plant new potatoes.

The fertilization of the greenhouse is also done autonomously. Over the entire greenhouse, sensors will be placed in the ground. These sensors will be measuring the humidity of the ground. When the sensors measure the humidity is below a certain point, they will give off a signal. This signal goes to a connected sprinkler system. The sprinkler system will give off a fixed amount of water to the crops. The sprinklers will turn off afterwards and 30 minutes later the sensors check the humidity again and if the humidity is still too low, this cycle will be done again. Once the sensors detect that the humidity is above a certain level again, the sprinkler system will not give off water to the crops.

### Goal

We want to make a fully autonomous greenhouse. We want to try and do this in the form of a model. In this model how the greenhouse is going to be will be described. But we also consider the user needs, the strange environment and the considerations of the society and enterprise. Though it is never possible to please every single party, we want to try to compromise as good as possible and get everyone to agree with our technology. But most important, we want the primary users to be happy with the result and we want them to be able to use it in the easiest way possible.

### Approach

We want to make our model by first looking at what parties are involved, what their requirements and preferences are, and if there are any constraints. Based on this information, a lot of research needs to be done on the process of farming potatoes, on the environment of Mars, on the already existing greenhouses and on (autonomous) farming robots. All of this information has to be displayed clearly. From here on we can collect hard data (numbers) on what potatoes need (for instance how many water per certain time). These hard data will be used in a program that will calculate how much kilograms of potatoes we will be able to gain in an amount of days in the greenhouse on Mars. Based on that, the size of the greenhouse can be determined.

## USE aspects

### Problem statement

There is a mission going on to send a group of people to Mars for colonization. In the first stage, this group consists of about 20-30 people. After this, every couple of years a new group is sent to Mars. When the first group arrives on Mars, it is important that there are enough resources to sustain. Since the plan is to stay on Mars for the rest of their lives, merely canned food will not suffice (it will run out fast). Thus food has to be tailed. As there is a limited amount of crew members in the first stage of this mission, it is favorable that as much work is done autonomously by rational agents. One of the most important things in order to survive, is the tailing and harvesting of food. In the beginning, this should be focused on nutritious food, which is easy to process. For grain, as an example, it takes a lot of work to make it into something eatable, which is not favorable at this stage. We chose to look into the tailing of potatoes. This process is not yet fully autonomous.

### Users

The users can be divided into three groups: primary, secondary and tertiary users. The primary users are the users that directly come into contact with the technology or who directly benefit from it. The secondary users usually do not directly use the technology, but benefit from it anyway, for instance, when they work on the development or design of the project. The tertiary users are mostly involved in purchasing the technology.

#### Primary users

Inhabitants of Mars

The direct users of our technology will of course be the inhabitants of Mars. The technology is designed especially for them, so that their life on Mars gets a bit easier. In the first stage of the Mars mission, all the inhabitants are going to be (mechanical) engineers, doctors, biologists, agriculture experts, physicists etc.. They are all going to be trained in advance to be able to handle all kinds of situations. They will be trained to use the technology that is sent up with them, including our technology. There are no experts for all the different kinds of technology though, which means all of the technology needs to be relatively easy to understand.

#### Secondary users

(Mechanical) engineers (on earth)

(Mechanical) engineers are needed in order to design and build the technology on earth.

Programmers (on earth)

Programmers are needed in order to make the process of potato tilt fully autonomous. They need to implement all of the research done by other scientists in order to create the perfect environment for potatoes to grow in.

Biologists (on earth)

Biologists need to do research on the environment on Mars and in and how growing potatoes is possible in that environment.

Agriculture experts (on earth)

Agriculture experts need to establish the perfect kind of potato to grow under these extreme circumstances, as well as they need to determine the perfect conditions for the potatoes in each stage of their growth.

Chemists (on earth)

Chemists need to determine how to change the soil and air so that the right nutrients are in there.

Physicists (on earth)

Physicists need to calculate how the gravity is going to effect the growth of the plants.

All secondary users described above profit from the technology in the sense that it will be their job to contribute to the development of the technology, so they will get paid to do their jobs.

#### Tertiary users

The buyers of the technology also play an important role of the process. These are the ones paying for everything after all. If they are not happy with the technology, they will not pay, which makes it very important to look at their requirements and preferences as well. The buyers of our technology will probably be either NASA or the government.

### User requirements

#### Primary users

• In the early stages of the Mars mission, people might not have the knowledge to make any complex modifications or repairs to the robot. This has to be taken into account when the robot is designed. It is important to make a design that is as simple as possible and make the programming as clear as possible. In this case, when something does break, it will hopefully be easy to fix.
• Another consideration will be the material of the greenhouse. Because the resources are limited, it is important that when the technology malfunctions, it can easily be made without using a lot of materials and specific tools. If certain materials or tools might be needed in order to fix the technology, it is important to know that in advance, so these can be taken with the crew to Mars.
• The technology needs to be easy to use. Preferably, the technology is fully autonomous so that the crew does not loose precious time while building up Mars.
• The technology needs to be efficient. It is important that in an environment, where there are so little resources like food, water, manure or energy is not wasted.

#### Secondary users

• Money for research.
• Enough time for research. A lot of the time researchers are set with very strict deadlines in which it is not always possible to deliver the quality of work needed for such a big project.

#### Tertiary users

• The technology needs to be as cheap as possible.
• The technology needs to have a long lifespan, so that it does not have to be re-purchased.

### Society

Apart from the users there are two important groups left to consider, Society and enterprise. The two aspects are intertwined for some part but we will consider them separately.

First off is Society. The entire world population falls under this category because sending a small group to Mars could have a huge impact on the rest of the civilization. The Mars project is a really expensive one where only around 20 people will reap the initial benefits and get sent to Mars. This could cause a lot of uproar about the project, because some people might feel like they are investing billions of euros into that small group of people. Initially that feeling would be true, before a second group of people are sent to Mars, the first group is solely benefitting from the autonomous farm. However the autonomous farm could (with some changes) be implemented on Earth which would increase the efficiency of growing crops and the technology would therefore not only be used on Mars. So shifting the investment from solely Mars to both Mars and Earth might ease a lot of people into becoming less opposed to the project.

The next group of people that could start to counter the autonomous farm are the farmers. Since the implementation of autonomous farms that are more efficient than farmers could cost a lot of jobs on Earth. Farmers are a really large group of people that would all be opposing something that could eliminate their job altogether. There are different ways to deal with the farmer opposition.

The first possibility is instead of replacing the farmer with the autonomous farm, the farmer actually uses the autonomous farm to do all the hard work for him and the farmer could still distribute and sell his potatoes. However, that could still mean that a lot of the farmers lose their jobs since the autonomous farm does everything up and until the potatoes are harvested and collected, so distribution would be the only job left to do. And on top of that, the autonomous farms are supposed to be more efficient than a human farmer, meaning there are less farms needed to feed the same population. All things considered, letting the farmers use the autonomous farms still will not convince them of the benefits of an autonomous farm.

Another approach is to build the autonomous farms, not in more crowded and stable civilizations, but instead in third world countries. Where a single autonomous farm could feed a lot of people, causing the local population to flourish and grow. A sudden growth in population in these countries could have major effects in very different directions. One of the possibilities is that the growth and the flourishment of the country could turn it into an equivalent of the modern western societies with a lot of wisdom and knowledge getting spread throughout the country and eventually leading to a crucial country for innovation on multiple fields. An outcome to wish for, but not guaranteed. On the other hand there is a possibility that the growth of the population causes nothing more than an increasing number of people needing food and water without any progress in education or wealth in the country. With the original reason for the Mars project in mind, which is a second planet for humans so that Earth does not get overpopulated, the second scenario would only be countering the initial motive of the project. Of course in the third world country one or the other outcome could happen, or anything in between, so there is no way to know how an autonomous farm would change the country.

In conclusion, the Mars project is one of those subjects where there is always a group of people that is either opposed to the project, or a group where the investment and innovation has a negative influence on. But every big revolution and innovation has its opposition, so the Mars project should not be canceled because you cannot make every person happy.

### Enterprise

For Enterprise it is all about the value. Whether it is worth it to invest into a project depends on the gain. Once the Martian project launches actual people to Mars, it will be worldwide news. Having your brand all over the project will give your company recognition over the entire globe. Every news report will either mention the investors or some other form of thanks for the investment. It might be the most reliable way to spread your brand across the world.

### RPC's

In this section the requirements, preferences and constraints of the technology will be elaborated. With the limited resources, energy, and manpower that are available, the technology has to be altered to such an extent that all the different scenarios are to be taken in account. This will prevent early wear on the robot and its surroundings.

##### Requirements

All requirements are essential for the sustainability for the robot.

• Versatile mobility

The robot that can do multiple things, namely; planting potatoes, harvesting them, eliminate all remains of the potato plant to stimulate better growth for the next cycle, transport the potatoes, and do this while moving in between the different potato plants without trampling them. With that all to be taken in account the robots movement has to be versatile and reliable. Versatility is playing a large role in the development of the current technology, since this also lowers the cost of the development of the robot. The latter is due to the fact that the same technology can be used multiple times. Making it more beneficial for the enterprises, making it look better in the eyes of society, and by reducing the costs, more can be invested in other things.

• Efficiency

The means to send materials to Mars are very expensive and thereby very limited. While on earth we will be able to use means like fossil fuels and such. This is not the case on mars. A limited field of solar panels will probably be the most reliable source of electricity. Therefore the whole potato harvest process has to be as efficient as possible. Since no speed records have to be set by the robot, one can think of more dramatic ways to increase its efficiency. An example is the use of wheel motors to bring friction down to a minimum.

• Ease of transport

A system for transport is of course not difficult to implement in a robot, especially not for a vegetable like potatoes. However, the robot has to be compact even when it has a container to store the potatoes temporary. When the robot exceeds a certain size, it will be difficult to maneuver between the potato plants with care in a relatively small area. Adding too much weight will destroy the loose earth ground, maybe even to such an extent that the potatoes plants cannot grow optimally. Water that is sprayed on top could also be inhibited to get absorbed. The last, but very important, thing is that when the robot is larger, its weight will increase. This makes it less efficient due to the extra weight and resistance it has to overcome to drive. So with this in mind a system that can only carry a few potatoes might not be disadvantageous.

• Ease of maintenance

With only a very limited amount of personnel that will be send to Mars, it is essential in the first years or even decades that the amount of required maintenance stays as low as possible. The people staying on Mars will have more important issues to worry about, so some compromises might have to be made to keep the maintenance that low. For instance, when extra error margins have to be computed for, the robot might get heavier and thus less efficient, these kind of things have to be taken into consideration when designing the robot. But when something breaks in the robot, it needs to be fixed or replaced relatively easily. When designing, it can be made with different modules, which can be replaced by reserves that are to be send to Mars.

##### Preferences
• A connection to Earth

One cannot expect from the people, that will be send to Mars, that when the robot does not behave like initially intended that they can fix it, that is, when the problem isn’t physical. So when part of the arm does not rotate for an instance, the habitants of the Mars base will probably have to fix it themselves. But the problem can be in the programming of the robot as well, when the robot cannot maneuver through the greenhouse for an instance, which might be due to difference in gravitational pull from the planets. When a connection is to be made between the robot and Earth, people like scientists, programmers and physicists can change things in the programming of the robot. A second benefit of this will be that the people on earth will be able to gather data from the robot. Like tesla is doing with their cars the data can be used to write a new program for the next iteration of the robot.

##### Constrains
• Size

In some views a very large robot might look like it is beneficial, but like mentioned earlier, to be able to sustain on Mars, everything has to be as efficient as possible. Often when size increases it tramples efficiency, especially when a robot has to maneuver in a relatively small greenhouse. Making the robot smaller does, however, come with some challenges, the robot needs to have some force, for instance to pull the potato out of the ground. When having finished that, it has to drive with the potato to the next potato plant and do the same thing again, with the extra weight of the potato it is carrying around.

### Scenario's

Linda

Linda is a women in her mid-forties. She has done two masters in computation science and she has a lot of experience in working with different languages of programming, but she is specialized in only one language. From a very young age, she was interested in outer space, so when the opportunity of going to Mars was there, she was the first one to apply. After lots of procedures and years of training, she made it to the final crew and she is now living amongst the first people on Mars.

On Mars, there is a problem with the technology of the autonomous greenhouse. It does not function as it is supposed to do. The system does not react to the sensors anymore. After the crew has checked everything that could be wrong, the conclusion is that the system needs to be rebooted. After the reboot certain parts of the code need to be reevaluated and possibly rewritten. Since Linda is the most experienced person on the crew, this will be her job. However, she finds out the language in which the code is written is not a language she is specialized in.

In order for Linda to solve this problem, it is important that the code is very well commended (in every step of the code it is explained what is happening in that code). It is also important that the code is a simple as possible. Since Linda has a lot of experience with coding, she will be able to understand simple scripts of different languages. Besides these requirements, it is very important that there is a possibility of contact with earth, especially with someone involved in writing the code. This person can help her with the more complicated parts of the code if needed.

Steven

Steven is a 32 year old man. He did a master in Biology and he has worked with a multidiscipline group on how to produce food on Mars for the people who are going to live there for two years now. He is now very familiar with all the do’s and don’ts when it comes to the cultivation of different crops on Mars.

On Mars, there is a problem with the nutrients in the soil. All the nutrients washed out of the soil thanks to a brief failure in the sprinkler system. Due to this the crops that are already growing in this greenhouse are going to die if the people don’t jump in.

After some checks of the values the sensors in the soil gave and a quick analysis of the soil it turns out that the people need to start fertilizing in the greenhouse so that the soil gets enough nutrients again to let the crops grow.

In order for Steven to solve this problem, it is important that there are enough sensors in the soil to get all the values that are needed to determine what type of fertilizer and how much of it is needed. It is also necessary that there is enough of all the types of fertilizer that are needed on Mars in stock. This way Steven can put together the needed fertilizer and can put in in the system so that it gets spread on time.

Harold

Harold is a 29 year old engineer. He did a master in Mechanical Engineering and he has been involved in the design of the greenhouse that will be put on Mars for a year now. He joined the group working on the greenhouse to know how it is going to be constructed and to lead the group who is going to build it on Mars.

On Mars, after the greenhouse has been built completely it is already operational for about two months now. All crops are planted and everything was going as planned, but then the sensors in the greenhouse showed that the levels of O2 and CO2 where changing rapidly. If the people on Mars do not jump in, all the O2 and CO2 will leak out of the greenhouse which causes all the crops in it will die, because of the hostile environment that is created in the greenhouse.

The crew goes on to check the greenhouse and its surroundings, to discover that one of the side panels was leaking gas out of the greenhouse. To repair this leaking side panel the team decides to temporarily seal the hole by screwing a small iron plate over it. After this temporary fix a new side panel is made to fit this one and it is placed over the old one from the outside. After attaching the outside-panel, the one on the inside is removed and the greenhouse works as it should again.

## Theory

### Conditions on Mars

#### Soil on Mars

Climate change on the planet Mars is discovered by detection of ground ice. The water layer is 10 - 40 cm thick and occurs in between latitudes of 30 degrees and 60 degrees. This means that water is available on Mars. Because of this carbonates can be produced. The exact substances will be discussed below.

##### Ions

A layer of ice has been found at a depth of 5 cm to 15 cm. In the soils around the Phoenix landing site calcium carbonate has been discovered (3-5 wt%) by scanning calorimetry. It showed an endothermic transition at around 725 degrees Celsius accompanied by the evolution of calcium carbonate and the soil had the ability to buffer pH against acid addition. Its formation in the past was due to interaction of the atmospheric CO2 and the water on particle surfaces. The Martian soil has been further inspected, where around 10mM of dissolved salts have been found out of which 0.4 – 0.6 % perchlorate (ClO4). The other negatively charged ions included small concentrations of bicarbonate, chloride and possibly sulfate. The cations detected (positively charged ions) included magnesium, sodium and small concentration of potassium and calcium. Besides this, an alkaline pH was measured of 7.7 (with a margin of 0.5). These findings included that the soil at Mars has changed for the past years due to the action liquid water. It has also been found out that there is a mechanism to place ice at the surface: clouds of ice crystals precipitated back to the surface and formed a daily basis. Even though ions have been found, nitrate is an important ion for the Martian soil to be able to grow crops. In the table under the section “composition Martian soil”, the concentrations of a few ions are given.

Nitrate should naturally be formed through the oxidation of atmospheric nitrogen and then accumulate on the surface. On Mars odd nitrogen (N and NO) can eventually be turned into NO2, which can form nitric acid. Nitrate, thus, has not yet been detected on Mars; there are only some possible detections. In table 3 multiple acid formations can be seen.

These results are bases on the Phoenix landing site in 2008. In 2014 a meteorite from Mars (EETA79001) has been investigated. When investigating this meteorite, it had been concluded that the soil composition was of Martian origin. The reason is that the detected substances within the meteorite are difficult to reconcile with terrestrial contamination. EETA79001 showed the presence of 0.6 ± 0.1 ppm ClO4-, 1.4 ± 0.1 ppm ClO3- and 16 ± 0.2 ppm NO3-. It has been said that because of the prenece of ClO3-, also ClO2- or ClO- should be present. This was then produced by Cl- and γ- and X-ray radiolysis of ClO4-.

Furthermore, an article was found comparing the findings on Mars and Antarctic Dry Valley (ADV) soils. The article contains tables including the concentrations of certain ions found in each place. It has been concluded that the salts of the Phoenix landing site and the salts of the meteorite are similar and that the ADV are also a good match. This also strengthens the argument that ADV is a good Mars analog environment of Earth. The occurring ions and concentrations of soil in ADV and on Mars are the same within one order of magnitude. ClO4- is three orders of magnitude bigger on Mars than on Earth. NO3- may be present on Mars, but this is at a level below the detection limit of the nitrate sensor. The soil of the meteorite and on the Phoenix landing site are alike, however, on average the concentrations in the meteorite are 4% of those on the Phoenix site and the concentration Ca+ is 16% bigger on the meteorite.

The last article includes information about the nitrogen cycle and the ratio of the loss of nitrogen isotopes. This mainly has to do with the air activity. Therefore, it is not relevant for the discussion about the ground. There was a good explanation, however, why it is difficult to detect nitrate on Mars. This explanation is given in the next section. A cause could be leaching. This means that the nitrate will be at a bigger depth in comparison to less soluble ions such as sulfates.

##### Nitrates on Mars

Some articles say the opposite of one another. Some state that nitrate has been detected on Mars and others do not. As nitrate is crucial for growing crops, it is essential to know whether this substance is actually present on Mars. The least recent articles conclude that nitrate is not present in the Martian ground. However, the most recent articles say there is. It could namely be the case that nitrate has not been discovered in 2001 but in between 2001 and 2013. Therefore it is the wisest choice to trust the most recent articles.

##### Composition Martian soil

Several investigations have taken place in order to find out what substances are located in the soil on Mars. First of all, the Phoenix Mars Lander WCL soil has been analyzed and the Mars meteorite EETA79001 sawdust. These type of soils have been compared to soil on the Antarctic Dry Valley and to a Mars simulant. This research is about being able to grow crops (potatoes) on Mars. In order to succeed into this, the easiest way would be to do experiments with the ground which is available on Earth. This is the reason why the Martian soil has been compared to soil on Earth. To be able to grow crops, the soil should be fertile and the amount of fertilizer must be known. What is in the fertilizer and the amount of is needed, is based on the composition of the soil and the needs of a certain crop (in this case the potato). In the table below, the concentration of ions is given for the meteorite and Phoenix Landing soil. Also the values for the EETA79001 meteorite are stated here. 1 gram of soil was added to 25 mL of DI water.
Table 1:

Ionic species Phoenix WCL mars (μM) EETA79001 Meteorite (μM)
Ca2+ 600 ± 300 1180 ± 1
Cl- 470 ± 90 13.2 ± 0.8
K+ 390 ± 80 1.8 ± 0.5
Mg2+ 3300 ± 1700 136 ± 1
Na+ 1400 ± 300 110 ± 1
NH4+ ND 62 ± 2
NO3- <1000 48.5 ± 2.5
SO42- 5400 ± 800 117 ± 5
ClO4- 2400 ± 500 1.02 ± 0.11

It has already been concluded that both of these soils are comparative for the Martian soil. It has not yet been decided which soil to take as a representative. Because of the fact that the numbers for what a potato needs are in grams, the amount of μM has to be converted into grams. In the table below, the amount of grams for each ion is given within one gram of soil.
Table 2:

Ionic species gram per gram soil (Phoenix) gram per gram soil (Meteorite)
Ca2+ 6,01E-04 1,18E-03
Cl- 4,17E-04 1,17E-05
K+ 3,81E-04 1,76E-06
Mg2+ 2,01E-03 8,26E-05
Na+ 8,05E-04 6,32E-05
NH4+ 0,00E+00 2,79E-05
NO3- 0,00E+00 7,52E-05
SO42- 1,30E-02 2,81E-04
ClO4- 5,97E-03 2,54E-06

In reality, each potato has access to a certain amount soil. The potato is placed in the ground at a depth of 10 cm. The space between the potatoes in the same row is 28 cm and the distance between potatoes in different rows is 75 cm. These numbers have been taken into account in order to calculate the volume of the soil that each potato is able to use. By taking the density of the Martian soil into account, the amount of grams for each ion can be calculated. The results are shown in the table below.
Table 3:

Ionic species gram total (Phoenix) gram total (Meteorite)
Ca2+ 1,26E+01 2,48E+01
Cl- 8,75E+00 2,46E-01
K+ 8,01E+00 3,69E-02
Mg2+ 4,21E+01 1,74E+00
Na+ 1,69E+01 1,33E+00
NH4+ 0,00E+00 5,87E-01
NO3- 0,00E+00 1,58E+00
SO42- 2,72E+02 5,90E+00
ClO4- 1,25E+02 5,33E-02

From the information above can be concluded that the Martian soil does not contain enough substances for the potatoes to grow on, therefore, the ground needs to be fertilized. The problem with fertilization on Mars is that there are no fertilizers. For that a solution has to be found. The 20 people who are going to Mars should produce the fertilizers themselves. This means that these people should use their own feces and pee to fertilize the ground.
Despite the fact that the Martian soil needs a fertilizer, the structure of the ground is also an important aspect. The potato plants should be able to absorb the needed water, oxygen and fertilizers. The ground on Mars has very sharp edges and water, oxygen and the fertilizers cannot reach deep enough to be picked up by the plants. This means that worms should be used. They can create holes in the ground structure to let more oxygen and water into the ground. It has been tested already by Wamelink that the worms do not die because of the sharp edges in the Martian soil. Furthermore, they have enough food from the ground and the dead plants.

#### Air composition

##### Air on Mars

The atmosphere of mars is composed of 96% CO2, 1,9% argon, 1,9% nitrogen, 0,15% oxygen and 0,05% CO. In the section “air composition”, the most important parts of the air composition will be discussed. These conditions are far from optimal for the growth of potatoes, and thus a non-hostile environment for the potatoes has to be created. This is going be to achieved by making a friendly environment in a greenhouse. What this friendly environment entails will be described below.

### Environment potato

#### Air composition

##### CO2

The CO2 level in air on Earth is about 401 ppm (0.00401 %) in which potatoes can grow well. But to get to know if potatoes can grow better in atmospheres with more CO2, tests using FACE rings are done. A FACE ring is a ring of pipes around a field of crops which emits CO2 so that the CO2 values around the crops rise and measurements with different levels of CO2 can be done. With the outcomes of these experiments it can be concluded that when the CO2 levels are raised to 660 ppm (0.00660 %) the Leaf Area will decrease, but the activity of the photosynthesis is raised and therefore the number of tubers will increase. While the number of tubers increased, the size and weight of the tubers remained the same as on Earth. With these values in mind it would be a good idea to make an atmosphere in the greenhouse that has a CO2 value of 660 ppm (0.00660 %). This makes the tuber yield higher and it thus creates more food from the same amount of plants.

##### O2

When potato plants are just planted, they will use a lot of O2. This means that the plants emit a lot of CO2. The plants need the oxygen to create the haulm, which is used for the photosynthesis. It is thus necessary that there is enough O2 available for the plants to grow at first, but when they grow to a certain extent, they will start to transfer CO2 to O2. Plants transform CO2 to O2 using the energy that they gain from light via photosynthesis, but when there is no light, the plants will start respiration. During respiration, plants do always the glucose, that is made by photosynthesis, to CO2, water and energy for the plant to keep living. But when there is not enough light, photosynthesis rate is lower than this respiration rate.

#### Gravitational influence on Mars

Some questions arose whether the gravity on Mars has a significant influence on the development of the potatoes. However, it is unknown yet whether it will. The plants do grow when there is a change of gravity, but their shape is not known. In the model we will assume that it has no influence, also because certain researchers including Wamelink think that it does not have an influence.

#### Ground humidity

After planting, it is important that the ground is kept moist until emergence. When the seed is kept moist, a rapid emergence is stimulated. However, it is crucial that the ground is not too wet. The seed also needs oxygen and it should not suffocate. (H.P. Beukema 1990)

In the stage after emergence it is important that the moisture of the ground does not exceed a soil water tension (SWT) of -0.3 MPa, for then the growth of the leafs during haulm growth is delayed. The soil water tension is the pressure needed by the plant to extract the water from the ground. It is very important that there is no over-irrigation in this stage of crop growth, as the roots of the plant then become superficial, whereas long developed roots are important for later stages. (H.P. Beukema 1990)

In the period between the time just before tuber growth initiation and during tuber growth initiation, the ideal moisture of the ground is strongly dependent of the breed. Averagely, when the ground is relatively dry, initiation of tuber growth will happen faster. However, the amount of tubers per plant will decrease. The latter also happens when the ground is too wet. (H.P. Beukema 1990)

For potatoes, the ideal water content in the ground is 65% ASW , when the tuber growth initiates (C.C. Shock 2007). ASW (available soil water) is the percentage of the maximum amount of water a plant can possibly consume from a certain volume of soil. It is important that the ground moisture is sufficient and not inconsistent as potatoes are relatively draught sensitive. When there is a dry period, tuber growth is delayed. This is not only caused directly by a lack of water, but also indirectly as the leafs become less efficient in absorbing light. Also the quality of the tubers will decrease due to draught, as cracks develop and the tuber will grow in an undesired shape. (H.P. Beukema 1990) It is not possible to know when the soil water content is 65% ASW. Therefore the soil water tension is taken into account. Sensors do exist for measuring the STW, which makes this a measure to work with. The ideal water tension for potatoes can range from -20 to -60 kPa depending on the ground used to grow on, climate and the irrigation system used. (C.C. Shock 2007)

#### Day length

For potatoes it is of great importance that a daily cycle is involved. The potato plant can sense when it is day or night with certain light receptors called phytochrome. These receptors are positioned in the leaves of the potato plant. (M. Rodriguez-Falcon 2006) When days are long, the growth of gibberellins, which are growth hormones in a plant, is stimulated and this inhibits the initiation of tuber growth. When the days are short, the initiation of tuber growth thus is stimulated as the amount of gibberellins is then reduced. (H.P. Beukema 1990) The length of the night in this process is rather important, as when a night break is simulated, the tuber growth is also inhibited (M. Rodriguez-Falcon 2006). Therefore long nights are essential for good tuber yield.

In the initiation of tuber growth in potato plants, it is for wild potatoes more important that the days are shorter than for modern cultivars. These modern cultivars show an initiation of tuber growth quite fast relatively independent of day length (a cycle is still important however, but there are bigger margins) (M. Rodriguez-Falcon 2006). In other words, every cultivar has its own critical photoperiod. Whereas some cultivars cannot grow tubers when the day length exceeds 12 hours (short cycle potatoes), others can grow tubers with a day length lasting until 18 hours (long cycle potatoes). There are even cultivars that are able to initiate tuber growth with a photoperiod of 24 hours. However, the latter is definitely not an ideal situation to grow tubers optimally.

In the figure below the tuber yield from different breeds is portrayed as function of the day length. This graph shows that short days are the best circumstances for both short and long cycle potatoes.

The day cycle is mostly an important aspect between emergence and the initiation of tuber growth. When tubers start to develop, the growth habit of the potato plant is not affected as much by day length anymore. From this moment mostly other factors will determine the tuber yield.(H.P. Beukema 1990)

The yield from a potato plant as function of the day length for different temperatures and different breeds (retrieved from H.P. Beukema 1990).

#### Temperature

Potato plants are also sensitive to temperatures. Different temperatures to different plants have been inspected. From these experiments turned out that for good potato yield a cooler soil is preferred. This means a temperature around 15 ⁰C - 18 ⁰C. In the same figure as referred to in the section about day length, it is illustrated that indeed for a temperature around the 18 ⁰C, higher yield is obtained than from warmer or cooler soils. (H.P. Beukema 1990)

It is found that when changing temperatures from 14 ⁰C at night to 24 ⁰C during the day can improve tuber yield by 25% with respect to a constant temperature of 18 ⁰C. This was tested with the breed “Delani”. For “Nordland” potatoes no significant changes were found. (S.M. Bennet 1991) This also illustrates how great the differences in properties are per cultivar.

Other tests were done with larger temperatures to different parts of the potato plant. It turned out that whatever part of the plant was heated, it had a negative effect on the overall tuber yield (C.M. Menzel 1984). This is not desired. This has also to do with the fact that when higher temperatures arise, more gibberellins develop in the plant, which are known to be disadvantageous for tuber growth. (E.E. Ewing 2010)

Just like for the length of the photoperiod, temperature as also the greatest effect on the potato plant during the period right before and during tuberization. It is thus important that temperatures in this time of the lifecycle of the plant are right. Especially for longer photoperiods. Then the temperature then has an even bigger effect. (H.P. Beukema 1990)

#### Light intensity

Irradiance is an important factor as, of course, plants grow with photosynthesis. The light intensity is one of the variables that influences the photosynthesis. Research shows that for higher intensities, the growth of tubers is stimulated and thus a higher yield is obtained. For lower intensities haulm growth is stimulated, but tuber yield is decreased which is an undesirable effect. It is also known that for higher temperatures, the light intensity has a more important role than for lower intensities. For example, potatoes growing in tropical climates need a higher light intensity to have a good yield than potatoes grown in northern Europe. The higher intensity is for both temperatures better, but when the intensity decreases, potato plants in Europe can still deliver a good yield, whereas in tropical climates the yield is disappointing. The fact that yield is decreased by lower light intensities has to do with an increase of growth hormones like gibberellins, of which we know that their effect on tuberization is negative. (H.P. Beukema 1990)

#### Manuring and nutrition

Potatoes need the following substances in order to develop themselves and grow.

Macronutrients

• Nitrogen (N)
• Phosphate (P)
• Potassium (K)
• Calcium (Ca)
• Magnesium (Mg)
• Sulfur (S)

Micronutrients

• Boron (B)
• Copper (Cu)
• Molybdenum (Mo)
• Iron (Fe)
• Manganese (Mn)
• Zinc (Zn)

In order to make sure the potatoes receive these substances, human feces and pee will be used, which will be discussed later on this page. The macronutrients are the most important nutrients, however, it can also be assumed that there will be enough of the Magnesium and Sulfur when the other nutrients are present in sufficient concentrations. The same goes for all of the micronutrients: when there is enough of the four left-over substances (N, P, K, Ca), there will also be enough of the micronutrients. The amount needed of each of these four important substances needs to be calculated. It could happen that there is a surplus of one of the ions. In that case, these have to be separated.
Each of these four important ions will be explained when there is a shortage of it for the potato plant.

Nitrogen: Nitrogen is important for the development of the plant and ensures a better yield. When there is too less of this substance the leaves will not fully grow and a light green discoloration will be seen on the leaves. But when there is too much of this substance, there occurs a delay in tuber initiation. This means that there would be a reduced yield. A surplus of nitrogen even can affect the tuber quality in a negative way.

Phosphate: Phosphate boosts the development of the tubers and it enlarges the amount of the tubers. It also ensures that the tubers are equally big. This substance actually improves everything that has to do with the tubers. When there is a shortage of it, the growth will be inhibited. A shortage of Phosphate can be noticed, because the cuttings will stand up and the color of the leave will be glazed and dark green. Furthermore, the edges of the leaves will curl up.

Potassium: This is also a very important substance. It improves the yield and its quality. Furthermore, the resistance against diseases will increase and the tuber magnitude will increase. When a shortage occurs, the plant growth will be oriented downwards and plants will not fully grow. A shortage can be noticed in the leaves as the underside is yellowing and the edges also curl up.

Calcium: A green and healthy leaf will exist because of Calcium. It ensures a better yield and quality. Furthermore, there will be less eruption during the storage of the tubers. A lack of Calcium causes internal scald and hollow tubers.

##### Amount of nutrition needed

As can be deducted from the section “composition Martian soil”, this soil does not contain a concentration of ions which is sufficient for growing potatoes. It is assumed as if the Martian soil contains nothing and that everything needs to be present in the fertilizer.
In the development and growth of the potato, there are certain stages where the potatoes needs more substances than in other stages. Furthermore, there is a difference between the amount of nutrients that the tubers need and that the haulms need. In the table below, the amount of the four important ions needed each day is given in mg.
Table 4:

0-15 15-30 30-45 45-60 60-75 75-90
P 3 13 46 47 47 16
N 28 130 255 179 187 57
K 27 120 358 160 194 65
Ca 8 43 113 4 2 1

Each row represents a certain substance and each column represents a certain stage in the cycle, each one covering a period of 15 days. The amounts given are needed every day for each plant.

### Farming equipment

#### Farming robots

Inside the greenhouse will be a farming robot. The robot we want does not yet exist, but is based on the literature study. All of the separate elements of our robot already exist, so there is good hope our robot will be available in the next couple of years. The farming robot inside the greenhouse will be able to plant, harvest, and fertilize the potatoes. The robot will look like a tractor, and it will thus do the planting, harvesting, and fertilizing in a similar way common tractors do it. The difference is that it will be able to do this autonomously. For this, the robot needs a very good navigation. Otherwise, it will manage some parts of the field twice and other parts not at all. Therefore, it can track where it has already been and where it has to go next. This can be done by using a camera detection system.

Given the size of the greenhouse, it is important that the robot has a small turning radius. Therefore, the robot will be able to move forward, backwards and sideward. To realize this, the robot has wheels in two different direction. It has four wheels pointing forwards (type one) and four wheels pointing sideward (type two). When the robot wants to move forwards, it will use the type one wheels in order to move. In this case the type two wheels are lifted off the ground so that they do not interfere with the movement. When the robot comes at the end of the field, it will put down the type two wheels and then lift the type one wheels so that it can move sideward to the next stroke of potatoes. This way, there is no need for extra space to turn the robot outside the potato field, which will be much more efficient.

For the fertilization of the greenhouse, two different options are regarded. The first option was fertilization through the sprinkler system. This seemed efficient at first sight, since there already is a sprinkler system for the water supplies. But when this option is inspected further, it turned out this might not be as efficient after all. This would have meant that there had to be a separate tank to connect the sprinkler system to, so that the system can switch between the two tanks. The idea here was to have two different pipes, that come together in one pipe. In this case, if only water is needed, the fertilization pipe would close and vice versa. But this would bring a very high risk for leakage. This will result in a lot of maintenance which is to be avoided as much as possible. Therefore a second option is chosen, which is implementing the fertilization into the robot. This is efficient, since the robot is already there. If the robot would only be used to plant and harvest potatoes, it would be a very expensive piece of machinery which will only be used about four times a year. Fertilization is a feature that has already been implemented in multiple tractors today. Using that feature, the robot will be of much more use.

Several articles were found on farming robotics. However, most of these articles focus on robots that are (partially) manually controlled. It has only been a couple of years since designing and experimenting with fully autonomous farming robots has begun. The hardest part in programming the robots is to make them able to see whether a fruit or vegetable is ripe, or even to tell apart a fruit or vegetable from the plant. Furthermore the robot needs to be very delicate in order for the plant not to be damaged.
In Hill, P. (2016) a robot is described that is used to farm tomatoes. In this case, color is used as a measure of how ripe the tomato is. When the color is of a certain tint, the robot picks it. Otherwise it waits until the tomato gets a darker shade. The robot, which is described, is a very small robot which climbs up the plant and cuts the stem of the tomato if it is ready. This technique is however very hard to apply in this case, since there is a focus on potatoes which grow under the ground. Though, there might be a way to use color underground if we introduce a light in our robot. Another downside to this technique is that it requires a lot of these small robots, which can turn out to be expensive. On the positive side, these robots are very precise and do not waste a lot of vegetables (once the programming is perfected). An adjustment can be made from looking at the color by looking for example at the size when it comes to potatoes.

## Assumptions

Since this model will be a very first draft, lots of assumptions and approximations need to be made. All of the made assumptions and approximations will be listed below, including an explanation of why these are valid. Throughout the project, as many approximations as possible will be eliminated, in order to make a more detailed model. The assumptions will be divided over a couple of subsections, such as the things that the robot is assumed to do, the composition of the ground and the growth of a potato. This in order to try and give a clear overview of all the simplifications in our model.

#### Users

It is important for the developers of the technology to keep in mind what tasks the users can perform. This way the developer can make sure that the technology can be maintained as well as needed and that the technology will have a long lifetime. For this, a couple of assumptions are made about the primary users.

• The primary users are trained in solving minor problems with all kinds of technology
• There will be (mechanical) engineers amongst the primary users with the skill of performing a bit more complicated maintenance
• The Martian crew will be able to contact earth at all times
• The crew members have minimal knowledge about potato plantation
• The crew members have minimal knowledge of the specific technology

#### Robot in greenhouse

At the moment, there are no fully autonomous farming robots available that fit in our project. There has been done research, though, on making tractors autonomous, so the robot in the greenhouse will be based on this research. Since this robot does not yet exist, a lot of assumptions need to be made. All of these assumptions are based on the literature research.

• The robot is able to plant potatoes
• The robot is able to harvest potatoes
• The robot is able to plow the ground
• The robot is able to move forward and backward
• The robot has a small turning radius
• The robot is able to detect where in the field it is
• The robot can track where in the field it has already been
• The robot can move forward and sideward, but it can only plant, harvest and plow in one direction

#### Ground on Mars

In order to grow potatoes on Mars we need to know what the soil of Mars contains regarding for instance nutrients. Though, there has been a lot of research about the soil on Mars and the exact composition is hard to find. In order to make the soil right for potato growth, it needs to be fertilized. Before this fertilization can be included in the model, a set of assumptions and basic boundary conditions needs to be made.

• The composition of the Martian soil is seen as equally distributed around the whole planet
• The values found in the literature do not include errors
• Only take into account that the potato is able to take up ions from above it (so 10 cm in depth to take up these substances)
• Take size 28/35 potatoes (so width between potatoes in same row 28 cm) if size 35/55 was taken than it should be 38 cm. It is better to have more potato plants in the field, therefore, size 28/35 has been taken.
• Each potato is able to absorb the nutrients from half the distance to the next potato.
• If there is shortage in ions in the ground this can be easily dealt with by putting the shortage in the fertilizer.
• Human feces and pee will be enough to fertilize the ground.
• Other crops will also be able to grow on Mars, therefore, the dead plants can also be used to fertilize the ground.

#### Greenhouse

• The greenhouse is a closed system (every feces or dead plants etc. will be reused)

## Model

In this section, a description of our model will be made. Our model consists of a program that calculates the amount of potatoes that can be farmed combined with a description of what we think the greenhouse should look like. Furthermore, a section will be included which describes vital information that is too difficult to implement into the model at this stage. However, there has been done research on these subjects and thus it will be described what needs to be considered in a next version of the model, and a way to possibly implement them in the future.

### Why modelling is necessary

Every ingenious invention needs to be tested before it is used and sold. The same goes for everything that goes to space. For example, the entire rocket needs to be tested, but sending the rocket to space just for testing is a huge waste of both time and money. Therefore, every section of the rocket is either tested or modelled or both. Modelling is a very efficient and cheap way to test your engineering invention. Also for our project modelling is necessary, since there is no way to test what the actual amount of potatoes is, that would be retrieved after harvesting. Because sending the entire greenhouse to Mars, simply for testing, would be a waste of both money and time.

Even though modelling is a very useful tool for physicists, for biologists it is a lot less accurate. Physicists rely on clear formulas and an understanding of the subject to clearly model whatever their problem might be. Biologists however need to deal with something called life. Modelling plants or animals is difficult because living things can always stray away from the formulas used to model them. Every plant and animal is unique and grows, feeds and festers in a different way. This makes accurately modelling the amount of potatoes gained from the entire greenhouse after the entire cycle of growth, practically impossible. But modelling can still be a useful tool. Even though it is not clear how the plant reacts on every shift in air composition or every drop of water. Through empirical research done at several places in the world, hence different weather conditions, a simple model can be constructed to give an estimate of the amount of potatoes harvested.

Another possibility for testing could be testing the greenhouse on a farm on Earth. It helps the knowledge of how the greenhouse works and what the flaws are, since it is supposed to be completely isolated with a regulated temperature, day-night cycle, fertilization, sprinkler and air composition. But once the potatoes are harvested, that specific amount does not give the full picture since the ground is not at all like that on Mars. However, it is still more accurate than modelling. To get to the optimal conditions, just like the model, a lot of iterations will be needed. When testing the actual greenhouse on Earth, it will take decades to find the optimal conditions. Furthermore, all of the different aspects of the greenhouse and the robot need to exist already. With the current assumptions that are made, the robot can do numerous things, that already exist, but are not implemented into an autonomous robot yet. That would need a lot of research before the robot has every ability necessary to fit into the greenhouse system.

The testing on Earth might be a more accurate representation of the harvested amount of potatoes, but the actual testing could only take place in the far future. On top of that, it would need decades of testing before the optimum is found. Therefore, modelling is the more efficient choice, it is cheaper and less time consuming. Even though the accuracy might be further off, plants are alive and will never be bound by formulas.

### Manuring

#### Fertilization

Human feces and pee could be used to fertilize the ground. These substances, however, contain bacteria and parasites which can cause diseases to humans when using it. When sterilizing the feces and pee, this will not be a problem anymore. Since Mars has temperatures that could have a value of -50°C or even -100°C, this can be perfectly sterilized. The sterilization can take place outside where the feces and pee are exposed to the cosmic radiation. This type of radiation will can kill bacteria and parasites together with the extreme low temperature. It also shows that a closed system will be used on Mars, meaning that everything will be reused. The ureum from the pee will be used for fertilizing the ground and the water will be used for drinking. The compost of dead plants will also be used as fertilizers.

##### How to realize the fertilization

Humans produce 150 grams of feces and 50 grams of pee every day. Within the 150 grams, 35% is dry and within the 50 grams of urine 65% is dry. In the table below, the percentage is given of the substances within the dry fractions of the feces and urine.

% in feces/day/human % in urine/day/human
P2O5 3 2.5
N 5 15
K2O 1 3
CaO 4 4.5

Now that these numbers are known, the amounts can be calculated in grams and the P, N, K and Ca should be calculated (which is done by molar ratios). Furthermore, these amounts should be multiplied by twenty as there will be twenty human beings on planet Mars. When these calculations have been carried out, the following table can be made. It represents the amount of the substances available in grams each day.

grams/day
P 20.83874
N 150.000
K 24.90424
Ca 50.92117

Now that the amounts of available and needed fertilizer are known, one can calculate when the fertilization should take place and how much should be used each time.
The number of potato plants is required to be known in order to do a calculation on how much of the fertilizer is needed each week. From the section above the surface area of the field, it is known that 2794.5 kg of potatoes will grow on the field. According to Wamelink, each potato plant has around 2 kg potatoes. The numbers of potato plants on the field is, therefore, 1397.
When the number of potato plants is known and the amount of the nutrients for each plant for each stage, the amount can be calculated of what all of the potato plants need for each stage. .
It has been decided to fertilize each week, therefore, at each stage fertilization will take place twice. When fertilizing each week, less fertilizer is required, which ensures that the fertilizer does not become too heavy for the potato plants. Because of the fact that in some stages the potato plants need more fertilization than on other stages, the amount of the fertilizer placed on top of the potato plants also differs. In the table below, the amounts needed for all the potato plants for seven days long is given in grams.

stage 0-15 days stage 15-30 days stage 30-45 days stage 45-60 days stage 60-75 days stage 75-90 days stage 90-105 days stage 105-120 days stage 120-135 days
P 29337 127127 449834 459613 459613 156464 68453 19558 0
N 273812 1271270 2493645 1750441 1828673 557403 176022 48895 9779
K 264033 1173480 3500882 1564640 1897126 635635 176022 39116 9779
Ca 78232 420497 1105027 39116 19558 9779 9779 9779 9779

A calculation has been carried out, how much of the fertilizer is needed every week. It will be assumed that for every stage two weeks of fertilization take place. In order to ensure that enough of the fertilizer is available, it will be assumed that the amount needed for one week equals the amount needed for eight days. In this way, there is a certain error propagation when the potato plants are not able to take up all of the ions available. While calculating the amounts of fertilizer necessary, it was noticed that there was a clear surplus of some ions. The amount of Potassium in the fertilizer ensured that there was a surplus of the rest of the nutrients (N, P, Ca). It could be harmful for the potato plants when there is too much nitrogen available. In this case, the surplus of nitrogen is not harmful for the potato plants, as the surplus is not too massive. When there is a surplus of the rest of the nutrients, it does not matter. It is simply not absorbed by the potato plants. In the table below, the amounts of the fertilizer needed for each week is given.

week 1 week 2 week 3 week 4 week 5 week 6 week 7 week 8 week 9 week 10 week 11 week 12
Days of produced fertilizer by each person 213 213 943 943 2812 2812 1257 1257 1524 1524 511 511 142 142 32 32 8 8

When the total amount of grams of the fertilizer is required for each week, the numbers in the table above should be multiplied with 200 grams, because this is the amount of feces and pee produced every day by each person. In the table below, the amount of grams per m2 is given in each stage.

stage 0-15 days stage 15-30 days stage 30-45 days stage 45-60 days stage 60-75 days stage 75-90 days stage 90-105 days stage 105-120 days stage 120-135 days
grams/m2 25,91240876 114,7201946 342,0924574 152,919708 185,4014599 62,16545012 17,2743917 3,892944039 0,97323601

The total mass in grams of the required fertilizer for each harvest can be easily calculated. The sum of the numbers mentioned above should be multiplied with the surface area of the field. After that, these numbers need to be multiplied by a factor 2, as each stage consists of two weeks of fertilization. Finally, when these numbers are summed, the total weight of the fertilizer for one yield of potatoes can be calculated. This number is equal to 2976,8 kg. The amount of fertilizer created by the 20 people on Mars in 180 days is equal to 720 kg. This indicated that people should already save their pee and feces before they are going to Mars. The amount of days they should start beforehand is equal to 564.2 days. To ensure that enough of the fertilizer is on Mars it has been decided to start saving pee and feces one year and seven months before departure.
This will only be applicable for the first twenty humans arriving on Mars, because after harvesting one field of potatoes, the dead plants can be used as fertilizer and the pee and feces produced on Mars. The dead plants are a better fertilizer than pee and feces, meaning that a smaller amount of it is required.

#### Surface area of the field of potatoes

Now, the size of the surface area, that is needed for the potatoes to grow on, is regarded. The graph retrieved from the simulation model (which is illustrated at the section “model explanation”), shows that the potatoes need around 180 days from being planted until they can be harvested. This means that within 180 days twenty people need to be able to live on potatoes. It has been assumed that each person eats potatoes for each meal. This means that every day three meals including potatoes will be consumed. Each man consumes 250 grams of potatoes for a meal and a female 200 grams. This indicates that on average, 225 grams of potatoes will be eaten by each person for each meal.
A calculation can be made how many grams of potatoes need to be harvested:

$\displaystyle{ \mathrm{225 \cdot 20 \cdot 180 \cdot 3 = 2430000 grams} }$

There is a possibility that some potatoes could obtain damage, for example. Furthermore, seeds are required for new plants for the next 180 days. This will also take some extra time. Taking both of these things into account, the weight of the potatoes is increased with 15%. This means that the new total mass of the potatoes will be 2794.5 kg. In the graph retrieved from the simulation model, it can be seen that around 17000 kg of potatoes are produced per hectare. From this knowledge, one is able to calculate the surface area of the field that is needed for the greenhouse:

$\displaystyle{ \mathrm{\frac{2794.5 \cdot 10000}{17000} \approx\ 1644 m&sup2;} }$

At some point in time new people will come to Mars. Enough food needs to be available for them as well. Now, we take in mind that people eat potatoes three times a day, because that are no other crops. In the future, when more people are coming, other crops can also grow on Mars. Therefore, people will eat potatoes once a day or even less. This indicates that there are at least enough potatoes available for 60 people. It shows, that the greenhouse for the potatoes is big enough for the upcoming years.
Because of the fact that Mars has plenty of space, there is enough room next to the potato greenhouse to grow other crops. One could imagine that after days of eating potatoes, the user is not satisfied anymore. When experiments can be done with other crops, such as wheat, other carbohydrate rich meals can be served on Mars. In this way, the users will be more satisfied. This is also true for the scientists, as they have been able to succeed in another sort of crop. The investigation of growing new crops is, however, not part of this project.

#### Seasons in the greenhouse

There has already been an experiment carried out by Wamelink to see whether potatoes can grow on a Martian soil in a green house. Potatoes indeed grew, however, they faced one main problem. The potatoes kept growing instead of dying. This created a lot of leaves and bad potatoes. The main cause according to Wamelink was that in the greenhouse there were no different seasons created for the plants to die. This induces that in our greenhouse, we should try to implement different seasons for an optimal result. These include varying the length of a day and the temperature smoothly.

### Greenhouse design

In this part of the page, several aspects concerning the design of the greenhouse will be discussed. These are the materials needed for the walls of the greenhouse, the lamps, the sensors in the greenhouse, and the sprinkler system.

#### Sensors

Sensors will be placed in the ground for measuring the soil water tension and the sensors will be placed in the air to measure the temperature. The sensor for the soil water tension is necessary to determine when the sprinkler system should become active. The sensors for the temperature are important to regulate the heat exchanger.

##### Moisture

The sensors for the soil water tension are called irrometers. In order to effectively measure the soil water tension, a shallow instrument should be used which is located 25 cm deep into the ground and also a deep instrument located at 46 cm deep. In the section “ground humidity”, some ideal values for the soil water tension are given. When the sensors measure a moisture of above those ideal values, the input in the model will make sure that the sprinklers are switched off. If the measured value describes a dryer ground than required, then the input in the model ensures that the sprinklers spray the water onto the potatoes. The sprinkler will be turned off after a set time, since it will take some time for the water to drain into the ground. Therefore it cannot be measured immediately. It will not be the case that the values for the soil water tension will be too low, as the ground is dry from itself and the sprinkler will be turned off in time.

##### Temperature

These sensors measure a certain value for the temperature. If the temperature value is too low, then the sensor will give a signal to the heat exchanger to heat the greenhouse up. If the value for the temperature at the sensor is in between ideal values, then the heat exchanger will stop heating. When the values for the temperature are too high, then the heat exchanger will make sure it cools down. When the ideal value is reached again, the heat exchanger will stop cooling. These sensors are located both in the ground at the height of the tuber growth and above the ground measuring the air temperature. This is due to the fact that air temperature can be different from that of the ground. In this way both ground and air temperature can be regulated separately. This is ideal as the heat capacity of ground is much higher than that of air. Later on this page is explained how the ground and air is heated in the section “heat exchanger”

On earth we have the special privilege that we have a relatively dense atmosphere and an active magnetosphere. In the magnetosphere, charged particles are affected and manipulated by the magnetic field of an astronomical object, in this case Earth. These charged particles are more useful for the habitants of Earth than one might think. For instance, the solar winds that occur on our sun can launch a cloud of charged particles, which is a plasma. This plasma consists of mostly electrons, protons and alpha particles. These are redirected by our magnetosphere and this creates the phenomenon of the northern lights. Mars however does not have the magnetosphere. because it is absent, the solar winds, charged particles and other cosmic radiation reach the ionosphere of Mars directly. The cosmic radiation then strips (air)particles from Mars, making a very rarefied atmosphere in comparison to the one on Earth. The cosmic radiation also reaches the surface of Mars. If the greenhouse would be made out of a transparent material, it would also reach the potato plants. When plants come in contact with the cosmic radiation, it can change the DNA of the plants with consequences that can differ from situation to situation. For instance, it can make them inedible, kill them off completely, or make them grow even faster. Some scientist are however not completely sure what will happen, because the plants that are on Earth now have been through millions of years of relatively low cosmic radiation. But to ensure that the plants will stay roughly the same, a completely sealed greenhouse will be build.

It is not possible to stop cosmic radiation, but what can be done is to break it up. When one breaks the particles up in even smaller particles they become less harmful. Materials that are good at breaking the particles up are materials that are as small as possible. The best are Hydrogen, Boron, and Nitrogen. Now it is very important that the materials can be used as a building block as well, the materials can of course be used as extra layers in between the building and structural layers. Only with efficiency in mind, for transport (the flight to Mars) and the building of course, the walls can best be build out of these materials as a whole or at least for the largest part. The boron nitride nanotube is for this purpose a very good option. It has the shielding capabilities needed while strong building capabilities remain. Especially the boron is very effective at breaking the neutrons, the nitrogen does not reach those standards but it also breaks the neutrons pretty effectively. Furthermore, one has to think about the thickness required for the walls. Because of the fact that the cosmic radiation is very strong, the walls need to be thick enough.

##### Construction material of the greenhouse

Besides the walls, an extra construction material is needed for the stability. There are some very fast winds, but that will not have much effect on the construction. The atmosphere is very thin on mars which means that it will only blow up the dust. Still, some columns are placed in between the walls and for that, carbon will be used. The other regarded materials were steel and aluminum. The costs strength and density of the materials are inspected.

Steel Aluminium Carbon
Tensile strength (Mpa) 400 150 4480
E-modulus (Gpa) 200 69 245
Density (kg/m3) 7800 2700 1800
Costs (euro/kg) 0.60 3.10 50

The factors “tensile strength” and “E-modulus” are both a measure for the strength of the material. As can be seen in the table, carbon has the highest strength. In this way, one can also use less columns than when using steel or aluminum. Carbon also has the lowest density, which makes it easier to build and also easier to transport to Mars. The costs of carbon are much higher, though. It could, however, be argued that when less columns are needed, less costs will be in the transportation aspect of the material. In this way, the extra costs could be erased. Furthermore, all of these materials have a good processability and recoverability. Steel and Aluminum are a bit better in this, however, carbon is still good enough to be used. So, the decision has been made to use carbon.

#### Lights in the greenhouse

As has been mentioned before, the experiment performed by Wamelink experienced some difficulties concerning the seasons in a greenhouse. These have not been implemented when they did the experiment. This was most likely that the potatoes were not as good as they hoped them to be. In the model for this USE-project, the day length and the temperature will be dealt with as a reflection for the seasons. According to Wamelink this should namely be enough representation of the seasons. By using lamps in the greenhouse, the day length can be adjusted. When the greenhouse has summer conditions, the lamps will be turned on for more hours than when the greenhouse has winter conditions. Based on several aspects explained below, LED lights have been chosen as lamps that are to be used.

Before, high pressure sodium lamps were used in greenhouses on Earth, however, when using LED lights 25% of the energy in the conversion of electricity into light is saved. Furthermore, 30% of the electricity can be saved when placing the LED lights in a smart way. So, a total of around 60% can be saved. This would also be helpful for the lamps on Mars; less solar panels are needed which will reduce the costs in transportation to Mars. Furthermore, LED lights hardly produce heat, which makes the greenhouse easier to regulate, as the model will only need to take care of the heat transportation in the heat exchanger. Another advantage is that the lights can be placed closer near the crops when needed. The optimum of the closeness of the lamps to the crops, differs for each crop. In this way, the distance from the LED light to the crop can be adjusted easily. In addition to this, the LEDs can also be placed on the sides of the potato plants, enhancing the photosynthesis. It is not clear currently where these LEDs should exactly be placed. Because a robot has to move through the potato plants, it has been decided to just place the LEDs against the ceiling. The different colors in light can also be realized with LEDs, as each phase of the potato plants needs a different light color as an optimum. There is still some research going on in this part, so it is not clear yet which colors to use. However, if we use LEDs, it is known that its color can be easily adjusted. Light colors also have two other advantages: it will enhance the plants’ resistance against diseases and it influences the opening and closing of the stomata.
The entire ceiling of the greenhouse will be filled up with these LED lights, so this means around 50-60 m2.

#### Sprinkler system

To ensure that the water will be equally distributed over the field of potato plants, a sprinkler system is used. As discussed in the section “sensors”, the sensors will be the ones that trigger the sprinkler system based on the soil water tension.
The sprinkler system will consist of several sprinklers and a pump to push the water out. To save energy costs, the sprinklers will be placed on the ceiling, meaning that less pressure is necessary, so the pump has to do less work. To distribute the water more evenly onto the plants, the sprinklers eject the water in 360°. A decision must be made between buying more sprinklers, which is cheaper or less, which makes the pump work harder. The space that the sprinklers need in the spacecraft to Mars is not much and the sprinklers themselves do not cost much. They only have to be bought once. The pump needs to work for a longer period and the energy needed will rise, meaning that more solar panels are needed. Solar panels are expensive and bigger than a few extra sprinklers, therefore is has been decided to take more sprinklers. Furthermore, if the model can be made more complex, only a few sprinklers can be switched on when a part of the field is too dry. In this way, the system will be more efficient and ensures that the other parts in the field will not get too much water. Furthermore, the width of the tubes in which the water flows must be decided. As the water pressure will not be high and less water is needed for each sprinkler, thin tubes can be chosen. This saves money and a bit of space in the aircraft. Also, a smaller nozzle will have to be used to let the water flow out as the pressure is low.
On earth, a sprinkler on the ceiling could spread water for 6 - 20 m2 on Earth. Because there could be an error in the sprinkler system, it would be too risky to assume that 20 m2 could be reached with one sprinkler. Therefore, a sprinkler in the greenhouse could cover 19 m2 as a maximum. Using the arguments mentioned above, sprinkler will be used covering 6 m2. This indicates that the number of sprinklers needed is 274 in total. The sprinklers will have to be evenly distributed to let each potato plant must the same amount of water. The number of sprinklers used in width and length will be equal to 137. The influence of the gravity must not be forgotten here. The pressure inside the tubes should already be higher than when placing this greenhouse on earth because the gravity on Mars is smaller. This extra amount of energy can easily be covered by the solar panels that are already placed for the greenhouse, meaning that no extra solar panels will have to be bought for the extra energy required.

#### Heat exchanger

Within this greenhouse, direct sunlight cannot be used for heating up the inside of the greenhouse. In addition, no window can be opened to cool down. Instead, a heat exchanger will be used to get the right temperature inside. By means of the sensors discussed above and the model, the heat exchanger will cool the greenhouse down or heat it up. Because of the fact that the system is closed, one should use cold and warm water to regulate the temperature. Because of the fact that the water on Mars is in the form of ice, this should at first be heated up until it turns into a liquid.
If the greenhouse should be heated, the water flowing through the greenhouse will be heated by the heat exchanger. This part of the system works the same as a heater in an ordinary home on Earth. Normally, an air-conditioning would be used for cooling the interior of the greenhouse, however, the system is a closed system unlike a house on Earth. Therefore, a solution has to be found to cool down the system.
It is possible to cool down the greenhouse by means of cold water streaming in tubes inside the greenhouse. One could argue that extra cooling might not be needed because on Mars itself it is cold enough. However, as the greenhouse is well isolated, it could happen that it needs to be cooled at a certain time. At daytime the temperature is higher than at night. The principle of the cooling mechanism would be the same principle as a heater, only this time the interior is cooled down. The electricity consumption for this system will be less than using an air-conditioning, because the condenser of an air-conditioning can easily reach a temperature of 50° and when cooling with water this will not be the case. The only problem with cooling by using water is that condensation can occur from the air in the interior on the tubes. This could cause the material to rust, however, one could also buy stainless materials, such as stainless steel, in order to not make this a problem.
Some tubes are also located in the ground. There are two types of tubes; the ones that have the function to heat the ground and the ones that lead water to heat the air. The temperature of the ground is at least as important as the air temperature. These temperatures can differ as explained in the section “sensors”. When the ground is too cold, warm water is redirected through the tubes and the ground heats up. When the ground is too warm, it can be cooled by the irrigation system. However, the ground should not become too moist, so if the humidity of the ground is high at the moment, then cold water can be redirected through the tubes. For the tubes leading water used for heating and cooling the air in the greenhouse, it is vital that they are isolated well in the ground.

#### Energy source: solar panels

When the autonomous greenhouse wants to work, an energy source is needed. This will be done by solar panels. The external irradiation between Mars and Earth is very different. For Earth this value equals 1361 W/m2 and for Mars this value equals 589 W/m2. This external irradiance on Earth is partially reflected, partially absorbed and there is some direct and diffused irradiance. Taking these things into account, 1000 W/m2 sunlight hits the surface of the Earth and, therefore, reaches the solar panels. For Mars it has been decided to take the same ratio. This means that for the solar panels 433 W/m2 can be used. However, on Mars there are less clouds, so the light intensity is a bit higher. On the other hand, more light will be reflected on Mars, causing the light intensity to drop. Some rough numbers can be considered here. Regarding the fact that Mars has more irradiance to pass through the atmosphere, the intensity will go up with 50 W/m2, however, on the other hand, more light is reflected, so the intensity will go down with 30 W/m2. This means that in total the light intensity on Mars will be 433+20 = 453 W/m2. This value is crucial to find out whether solar panels are even possible. The conclusion is that around twice as much solar panels should be used than on Earth. It is not exactly known how much solar panels are required as the energy consumption of the machinery in the greenhouse is not clear. The roof of the greenhouse is not flat, so it is not an ideal surface to place the solar panels. However, on Mars there is enough space to build a solar panel farm which generates plenty power for the greenhouse. The optimum angle for the solar panels can be calculated by means of the angle of the rotation axis. The angle of the rotation axis on Earth equals 23.5° and the optimal angle for the solar panels equals 35°. The angle of the rotation axis on Mars is 25°. This does not differ much with the angle on Earth. It has been concluded that the optimum angle of the solar panels will be 35° on mars as well. Even if this is not the exact optimum, the efficiency will drop with 1% when the difference is only 5°. In this case, the drop in efficiency is negligible. Therefore, the angle under which the solar panels will be placed equals 35°.

When looking at the costs of solar panels, it is quite an investment for the companies. It is however known that within 20 years one could benefit from choosing solar panels. Currently, a group of twenty people are going to look whether life on Mars is possible. This means that when it is possible, even more people could go to Mars and decide to live there. The greenhouse, therefore, needs to last for at least 20 years. Furthermore, another way to get electricity is difficult to implement on mars. One could think of wind mills; however, those are way too big to install on Mars. The twenty users on Mars will have less difficulty with implementing solar panels than building wind mills. Furthermore, the gravity will have a big influence in it, because the blades should be turning. It is also not known whether these wind mills would produce enough energy. For solar panels one could calculate the light intensity and amount of solar panels easily to find out whether this would give enough electricity.

#### Storage

After harvesting, the potatoes have to be stored. The way how they are stored depends on whether they will be consumed or they will be used for planting. At first all potatoes need to be cured. This is a process where small damages in the skin are recovered and the skin of the potato thickens. This is done under a temperature of 15 ⁰C. The curing process takes between one and two weeks. After that the still damaged and infected potatoes need to be sorted from healthy potatoes. Due to the curing period, infected potatoes are easier to recognize by symptoms. Potatoes showing symptoms of disease are destroyed. Potatoes that are damaged in such way that the curing process is hopeless, can still be consumed. However, they cannot be stored due to too much evaporation. After curing the temperature of the storage is ideally kept at 3 ⁰C – 4 ⁰C. Under this temperature the tubers are conserved the longest. It is also important that the tubers are not stored in large quantities on one pile. They need ventilation as respiration occurs within the tubers. In this process glucose is, together with oxygen, converted into CO2, water and heat. If the potatoes are piled up too much, the stack heats up and the maximum storage time is decreased. The ventilation is needed to have the correct amount of oxygen in the room and also to keep the potatoes at the right temperature. It is crucial that the ventilation is not too abundant, as evaporation then is a too great factor. That would dry out the potatoes. Normally potatoes are stored in the dark. When they come in contact with light, they turn green and that is not desired. However, for seed potatoes it is a good option to store them in diffuse light. The crops growing from the seeds stored in this condition are healthier than the ones stored in the dark. When the potatoes create chlorophyll in these conditions, it is not potent, as they are only used as seeds.

### Model explanation

Potato growth in Mars climate (ground, temp., rain and vapor pressure from Mars)
Potato growth in ideal climate (ground, temp. and vapor pressure from Earth)

Within the model, one can change a number of things and, like explained in different parts of this page, the growth of the potato reacts to all of these in a different way. To model the numerous variables, some assumptions have to be made. However, this is because of the complexity of modeling a non-numerical process, or simply because of the lack of knowledge. Not all processes are examined and mapped out in a numerical way so it cannot always be modelled. Starting with the growth of the plant itself, the growth will vary from plant to plant and this has to be taken in to account, since it will affect things like the water absorption. That absorption will affect the total needed water supply. Water absorption differs between the different phases of the plant growth, and at different times it will absorb more or less water. This is because in the different phases the assimilation rate varies. A basic assumption is that when there is a higher assimilation rate, the needed water is also higher. A specific range of soil water content that can be controlled by a sprinkler system, by hooking the sprinkler system to a simple controller that measures the soil water content with a few sensors spread out over the greenhouse to get a good approximation of the mean water content. The water retention and hydraulic conductivity of the soil itself are other variables that control the water management. This all creates a large loop, making up the water part of the model. There are a lot more variables that are taken into account. Like discussed above, the properties of the ground are an important factor. Also the properties of how it interacts with water are important. The nutrient values are as well an aspect to take into account, affecting the model in a different way, since the regeneration value of the nutrition in the ground is also important. When the plants use all the nutrition, there has to be a way to refeed the ground so it matches the needs of the plant. Since the greenhouse on Mars will have a way to distribute fertilizer, the refeed will be almost instant. This is necessary since the ground on Mars has no organic contents, so the Martian ground will not refeed the soil by itself. The different variables used by the model are in data files, these files contain different sets of data, ranging over a certain time to keep somewhat realistic variable circumstances. The weather here on earth, of course, varies a lot, but in the greenhouse it can be controlled in a very precise way to achieve the best growth-time, largest potatoes, and most nutritious potatoes.

### Explanation results

#### Nutrition

Like explained above, the potatoes need certain nitrates that are in the Earths organic ground by nature. This is modelled with 3 different nutrients, these can be added with the fertilization equipment of the autonomous tractor. The three nutrients are: Nitrogen (N), Phosphorus (P) and Potassium (K). These values can range from 0 up to 100 kg/ha and do not depend on each other. When experimenting with the different values a very easy pattern is to be found, the potatoes grow in a linear pattern parallel with the amount of nitrates that are in the ground as long as there is balance between the different nitrates. Namely when one value of the nitrates approaches zero. Making for a very easy conclusion for this part of the optimal values of the potato growth. In reality it will be a bit different of course, the 100 kg/ha does not exceed the top limit of the potato, because if so, the potato growth will approach zero.

#### Martian ground

The variables concerning the ground are expressed in the model as ranges. The model uses the values for the water retention and hydraulic conductivity. The calculations are done in between the most extreme values that are found or calculated on Mars. This gives a first indication in what has to be added to increase the growth of the plants and what has to be done to increase the water flow.

## Conclusion

The people on Mars need a self-sustaining food supply in order to survive indefinitely on Mars. For that reason, a design has been made for an autonomous potato farm. With the help of a program, it could be determined whether the autonomous farm was feasible. Inserting the found optimal values a good estimate of the greenhouse was made. Using the feces of the humans on Mars, the greenhouse is able to sustain itself solely on resources on Mars. Recycling left-over potato plants for the next batch. This provides a self-sufficient greenhouse that can grow potatoes anywhere on Mars, or even anywhere on Earth.

## Discussion

The entire autonomous robot is based on assumptions made about the state of the art in farming robotics. But there is not yet any robot that has all the functions that are needed for the autonomous greenhouse. Even though there is a robot for every separate function, the robot that does all the actions needs to be designed. This could be one of the bigger problems since combining that many abilities in a single robot is challenging.

Another major part of our design is based on the program. The values that are currently implemented give an estimate on how much kilogram will be harvested per hectare after a certain time in days. If the estimate is far off, the people on Mars might have way too few potatoes to feed everyone. Being suspicious about the program is not unreasonable since the growth of a potato is a biological system. Simulation of a biological system is never completely accurate, there are way too many miniscule factors that could influence the batch of potatoes after the entire growth cycle. So the program that was used has room for improvement. Multiple factors that the greenhouse regulates are not considered in the program. With the current program it is impossible to find out what adjusting these factors will do to the potatoes. But implementing these factors is hard since the biology does not follow clear functions and graphs. Someone running the same values twice is most likely to find two different outcomes. The program runs on empirical data, meaning if the data is not collected, there is no way to implement them into the program.

Another problem is the fact that every potato breed is different. Before sending the greenhouse to Mars, all the knowledge about properties of a certain breed have to be obtained. It is also important to know which breed is the best for circumstances on Martian ground. The required values are discussed above, but what these exact values are has still to be researched.

## Literature search

To start of our project we are doing a literature research. We divided the literature search into the following areas:

1) The conditions on mars (including the effects on aggro culture due to these conditions)

2) Robots and machinery that are already used in outher space

3) Farming robots that are already being used on earth

For all of these areas papers will be searched and a summary of the overall findings will be given with refrences to the found papers.

travel time

- B.G. Drake, “Human Exploration of Mars Design Reference Architecture 5.0” (2009) https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20090012109.pdf

Mars conditions

- Wei Luo (2011) Estimating hydraulic conductivity for the Martian subsurface based on drainage patterns — A casestudy in the Mare Tyrrhenum Quadrangle retrieved from: https://digitalcommons.unl.edu/cgi/viewcontent.cgi?article=1268&context=geosciencefacpub

- Boersma, H. (2012). Waarom gebruiken we geen mensenpoep als mest? - Kijkmagazine. Retrieved from: https://www.kijkmagazine.nl/nieuws/waarom-gebruiken-we-geen-mensenpoep-als-mest/

- C. Leovy. “Weather and climate on Mars” (2001)

- F. Gifford Jr. “The Surface-Temperature Climate of Mars.” (1955)

- S. R. Lewis, et al. “A climate database for Mars” (1999)

- G.W. Wieger Wamelink, et al. “Can plants grow on mars and the Moon: A growth experiment on Mars and Moon soil simulants” (2014)

- M. Nelson, et al. “Integration of lessons from recent research for “Earth to Mars” life support systems (2006)

- S. Silverstone, et al. “Development and research program for a soil-based bioregenerative agriculture system to feed a four person crew at a mars base” (2003)

- Journal of Climate. Jun2001, Vol. 14 Issue 11, p2430. 13p. 23 Graphs..... no access

- Agriculture, Ecosystems & Environment. Feb2018, Vol. 254, p99-110. 12p....... no access

- Climate Research. 2009, Vol. 39 Issue 1, p47-59. 13p. 7 Charts, 5 Graphs, 1 Map..... no access

- Canadian Journal of Forest Research. Oct2010, Vol. 40 Issue 10, p2036-2048. 12p. 1 Chart, 1 Graph..... no accesss

- Global Change Biology. Jan2007, Vol. 13 Issue 1, p169-183. 15p. 1 Diagram, 7 Graphs, 2 Maps..... (no access)

- James I. L. Morison,Michael D. "plant growth and climate change" (2006)..... (no access)

- R. RötterS.C. van de Geijn. "Climate Change Effects on Plant Growth, Crop Yield and Livestock". pp 651–681 (1999).... (no access)

- Angela T. Moles. "Global patterns in plant height"- Journal of ecology. (2009).... access

Potato

- Adam H. Sparks. "Climate change may have limited effect on global risk of potato late blight" - Global change biology (2014).... (no access)

- Bakhtiyor Pulatov. "Modeling climate change impact on potato crop phenology, and risk of frost damage and heat stress in northern Europe"- Agricultural and Forest Meteorology. Pages 281-292. (2015).... (no access)

- D.T. Westermann “Nutritional requirements of potatoes” (2005)

- Van Ittersum, M.K. & Scholte, K. Potato Res (1992) 35: 365. https://doi-org.dianus.libr.tue.nl/10.1007/BF02357593

- H.P. Beukema, D.E. van der Zaag "Introduction to potato production" (1990)

- P.L. Kooman et al. "Effects of climate on different potato genotypes 2. Dry matter allocation and duration of the growth cycle" (1996)

- C.C. Shock, A.B. Pereira, E.P. Eldredge "Irrigation Best Management Practices for Potato" (2007)

- S. M. Bennett, T. W. Tibbitts, W. Cao. “Diurnal temperature fluctuation effects on potatoes grown with 12 hr photoperiods” (1991)

- E.E. Ewing, P.C. Struijk “Tuber Formation in Potato: Induction, Initiation, and Growth” (2010)

- C.M. Menzel “Tuberization in potato at high temperatures: interaction between temperature and irradiance” (1985)

- E.E. Ewing, P.F. Wareing, “Shoot, Stolon, and Tuber Formation on Potato (Solanum tuberosum L.) Cuttings in Response to Photoperiod” (1978)

- M. Rodríguez-Falcón, J. Bou, S. Prat “Seasonal Control of Tuberization in Potato: Conserved Elements with the Flowering Response” (2006)

- S. Behjati, R. Choukan, et al. “The evaluation of yield and effective characteristics on yield of promising potato clones” (2013)

Existing farming technology

- Tillet, N. (2003). Robots on the farm. The industrial robot, 30(5), 396. Retrieved February 23, 2018, from [1]

- Giles, F. (2017). Agricultural robots no longer science fiction. Florida grower, 110(2), 12-13. Retrieved February 23, 2018, from [2]

- Hill, P. (2016). 9 robots designed to enhance the farm workforce. Farmers Weekly, 165(3), 66-70,72,75. Retrieved from [3]

- Robert Bogue, (2013) "Can robots help to feed the world?", Industrial Robot: An International Journal, Vol. 40 Issue: 1, pp.4-9, [4]

- Victor Bloch, Avital Bechar, Amir Degani, (2017) "Development of an environment characterization methodology for optimal design of an agricultural robot", Industrial Robot: An International Journal, Vol. 44 Issue: 1, pp.94-103, [5]

- Bac, C. W., van Henten, E. J., Hemming, J. and Edan, Y. (2014), Harvesting Robots for High-value Crops: State-of-the-art Review and Challenges Ahead. J. Field Robotics, 31: 888–911. doi:10.1002/rob.21525

Robots and machinery used in outer space

- Rećko, M., Tołstoj-Sienkiewicz, J., & Turycz, P. (2017). Versatile soil sampling system capable of collecting, transporting, storing and preliminary onboard analysis for mars rover analogue10.4028/www.scientific.net/SSP.260.59 [Robotic arm for Mars Rover] Link to article

- Czaplicki, P., Recko, M., & Tolstoj-Sienkiewicz, J. (2016). Robotic arm control system for mars rover analogue. Paper presented at the 2016 21st International Conference on Methods and Models in Automation and Robotics, MMAR 2016, 1122-1126. 10.1109/MMAR.2016.7575295 [Soil sampling] Link to article

- Sakib, N., Ahmed, Z., Farayez, A., & Kabir, M. H. (2017). An approach to build simplified semi-autonomous mars rover. Paper presented at the IEEE Region 10 Annual International Conference, Proceedings/TENCON, 3502-3505. 10.1109/TENCON.2016.7848707 [Making a Mars Rover semi-automatic] Link to article

- Wong, C., Yang, E., Yan, X. -., & Gu, D. (2017). Adaptive and intelligent navigation of autonomous planetary rovers-A survey. Paper presented at the 2017 NASA/ESA Conference on Adaptive Hardware and Systems, AHS 2017, 237-244. 10.1109/AHS.2017.8046384 [Intelligent navigation] Link to article

- Parness, A., Abcouwer, N., Fuller, C., Wiltsie, N., Nash, J., & Kennedy, B. (2017). LEMUR 3: A limbed climbing robot for extreme terrain mobility in space. Paper presented at the Proceedings - IEEE International Conference on Robotics and Automation, 5467-5473. 10.1109/ICRA.2017.7989643 [Robot with climbing arms (maybe possible to use for planting and harvesting] Link to article

- Garrido, S., Moreno, L., Martín, F., & Álvarez, D. (2017). Fast marching subjected to a vector field–path planning method for mars rovers. Expert Systems with Applications, 78, 334-346. 10.1016/j.eswa.2017.02.019 [Vector field-path planning] Link to article

Soil on Mars

- Boynton, W.V. , D. W. Ming, S. P. Kounaves, S. M. M. Young,† R. E. Arvidson, M. H. Hecht, J. Hoffman, P. B. Niles, D. K. Hamara, R. C. Quinn, P. H. Smith, B. Sutter, D. C. Catling, R. V. Morris. (2009). Evidence for Calcium Carbonate at the Mars Phoenix Landing Site – Science

- Hecht M. H. , S. P. Kounaves, R. C. Quinn, S. J. West, S. M. M. Young,† D. W. Ming, D. C. Catling, B. C. Clark, W. V. Boynton, J. Hoffman, L. P. DeFlores, K. Gospodinova, J. Kapit, P. H. Smith. (2009). Detection of Perchlorate and the Soluble Chemistry of Martian Soil at the Phoenix Lander Site – Science.

- Kounaves S. P., Brandi L. Carrier, Glen D. O’Neil, Shannon T. Stroble, Mark W. Claire. (2014). Evidence of martian perchlorate, chlorate, and nitrate in Mars meteorite EETA79001: Implications for oxidants and organics – Icarus.

- Manning C. V. , Christopher P. McKay, Kevin J. Zahnle. (2008) The nitrogen cycle on Mars: Impact decomposition of near-surface nitrates as a source for a nitrogen steady state – Icarus.

- Mustard J.F. , Christopher D. Cooper & Moses K. Rifkin. (2001). Evidence for recent climate change on Mars from the identication of youthful near-surface ground ice – letters to nature.

- Smith M. L., Mark W. Claire, David C. Catling, Kevin J. Zahnle. (2014). The formation of sulfate, nitrate and perchlorate salts in the martian atmosphere – Icarus.

- Stroble S. P., Kyle M. McElhoney, Samuel P. Kounaves. (2013). Comparison of the Phoenix Mars Lander WCL soil analyses with Antarctic Dry Valley soils, Mars meteorite EETA79001 sawdust, and a Mars simulant – Icarus.

Greenhouse

- Marcelis, L. (2014). Met LED-verlichting energieverbruik glastuinbouw halveren - WUR. Retrieved from: https://www.wur.nl/nl/nieuws/Met-ledverlichting-energieverbruik-glastuinbouw-halveren.htm

- Zonnestraling en zonnepanelen - Het weer in Haaksbergen. Retrieved from: http://www.weerstationhaaksbergen.nl/weather/index.php/Weblog/zonnestraling-en-zonnepanelen.html

## New part: robot

### Introduction

For the year 2032, the Mars One organization is planning to land the first humans on Mars and establish a permanent human colony there. The first humans will start to build an empire by building a base where people are able to live. After the first base is built, it will be developed further and further. In the end, people on Mars should be able to live there without any help from planet earth. Mars should become a self-sustainable environment.

Before we get there, a lot needs to happen. Right now, Mars does not have an atmosphere, so there is hardly any oxygen in the air. All of the water present on Mars is in the form of ice and lies far beneath the surface. Right now, no one is able to live there. Nothing is able to grow there and that needs to change.

In this project, we want to get a step closer to life on Mars. We want to do this by taking a look at the implementation of food production for the planet. We chose to focus on the plantation of potatoes, since potatoes are very nutrient and there has already been done some research on potato plantation on Mars

In an ideal situation the process would be completely or at least partially autonomic. In an autonomic greenhouse, a place where potatoes can be processed without human interaction, a robot harvests and plants the potatoes in an autonomous manner. In this project the focus will be on the robot design and the complications that may come with designing the machine. An analysis on the different user aspects will be given and the consequences different decisions might have.

The robot has to meet certain requirements without compromising on different areas. In an ideal situation the robot would not need any checkups or maintenance and such. Knowing that this is not achievable, different solutions have to be found. Size, speed and functionality are all factors that have to be taken into account when designing the robot. How this all affects the different users, will be discussed below.

### Goal

The goal of this paper is to give a clear insight in what has to be thought of when designing a robot. This does not only include the technical aspects. An analysis of different consequences of the design will be given for the different groups that will work on or with this robot. although it is never possible to please every single party, we want to try to compromise as good as possible and get everyone to agree with our technology. But most important, we want the primary users to be happy with the result and we want them to be able to use it in the easiest way possible.

### User aspects

#### Primary users

The primary users are the users which are going to live on Mars. This group includes 20 specialized people. There are for instance doctors, mechanical engineers, biologists, physicists and agriculture experts. It indicates that most of these people already have quite some prior knowledge on technology and agriculture. In this part of the report, a clear elaboration will be given on the relation between the primary users and the farming robot.

Before even moving to Mars, the robot and the users will already be in contact with one another. Because of the fact that transporting between Mars and Earth will cost a lot of money and time, the primary users should have enough knowledge to assemble the robot on Mars.
This knowledge will be provided to them on Earth, by means of showing the assembling of the robot and providing them information with what they should do when something goes wrong on Mars. One might think that making extra robots of this kind for practice on Earth costs money which is unnecessary, however, this project might also be performed in third world countries. An explanation for this can be seen in the part “society” on this page.

Now, that people are familiar with assembling such a robot, the separate parts of the robot can be sent to Mars together with the people. Because of the fact that the mechanical engineers have the most knowledge of the robots, they should lead the project of assembling the robot. The prior knowledge of the assembling of the robot is also provided to the rest of the group in the case that the mechanical engineers die, for example. For the assembling of the robot, a booklet will be provided with a step by step instruction on how to assemble it.
Considering the fact that the robots need to be assembled, the assembling itself should be going smoothly. It should be easy to connect the parts and they should not be too heavy, as it will be connected manually. Just like Ikea furniture.

Once the robot is finished, it should be connected wirelessly to the software, making the robot autonomous. This will also be provided in a booklet, and it is best to let a programmer do this job. When something does not work, contact could be made with Earth, however, it should not be the case that people from Earth should come every time, because that costs time and money. This contact cannot go by means of a telephone conversation, as this could take up to 21 minutes before the one actually hears what the person on the other side said (Startpagina, 2012). Therefore, the contact will go via mail.

Because of the fact that the primary users have already enough tasks to be done on Mars for themselves, it should not take them too much time to keep themselves busy with the robot. This is also a reason why the robot has been made autonomous. A scheme will be made amongst the users who will perform the simple tests and the cleaning at which day. The procedures are described in the subsection "Maintenance" within the section "Design" on this page.

Despite the fact that the users come in touch with the robot by checking and cleaning it, the robot might could also make noise while doing its job. This could be disturbing for the users if they are doing their job for which they were supposed to go to Mars. Instead of lowering the noise of the robot, the walls will be insulated well enough to ensure that the users will not be bothered by the robot while doing their job.

At last, the humans do also interact with the robot when it comes to storage. They have to store the potatoes, when the robot has harvested them. The robot should not be too big, as it will then need a lot of space in a room of limited size. One could say that there is enough space on Mars, however, building extra big parts because of the storage of a big robot will cost too much unnecessary money. The people will also need enough space to move around and store their own equipment and have their living space. In this way, the robot should not be too big, so it can easily be stored.

#### Secondary users

The robot will be made by the secondary user. There are, of course, multiple engineers that will be going to mars and they have to be able to support the robot if necessary, like it was explained in the primary robot chapter. But building and designing the parts of the robot is not a task for them and will take months if not years to complete. It is important that the robot is a help for the primary users and not a burden. The secondary users serve the primary users in a sense that all the design related decisions that are being made are to make the life of the primary users as convenient as possible. They have to do this all while communicating a lot with the tertiary users, who are of course responsible for the finances that are available for the construction of the robot.

The secondary users have to design the robot such that the robot satisfies all the requirements that are listed in the section “user needs” below. But they cannot keep throwing money at problems, creative solutions have to be found to meet certain requirements. A robot that is not too large is ideal for keeping the costs down, since the transport of the robot is easier, as storing the robot in the spaceship will take less space. The fuel that the robot needs is kept to a minimum when the robot is smaller, since a lower weight and mass moment of inertia means that less work is needed in the motors to move the robot.

They also have to make sure that the inhabitants of Mars can have a good and healthy diet by increasing the variety of foods. Normally, astronauts eat special food which is very dense in nutrients, which is not sustainable. Adding different foods will solve that problem. Using the robot to achieve this makes the inhabitants on Mars more or less self-maintaining.

Thinking like this will eliminate some problems the engineers have to deal with, but it will create some new problems as well. A smaller robot with less power and available force means it is less capable to transport multiple potatoes at once, meaning that when the robot has to make more trips to achieve the same result as a bigger robot. That might not be as efficient as was thought before. Using new available technology, and working together, gives more resources to achieve goals that were first not reachable.

Besides the extra money and resources for the development of the robot, the development of the robot will be improved if the secondary users work together. The robot might not be cost-efficient enough right now to use here on earth. However, when the robot is simply already developed for Mars, modifications can be made to use this new technology on earth, giving another reason to invest. These secondary users can make more money with these inventions, also giving a boost in future technological developments.

The secondary users will develop the robot with the Mars mission in mind, meaning that contact has to be close with the primary users. The primary users will have some expectancies about the different functionalities of the robot but also in how one wants to interact with the robot. With this in mind the designers will start to design the robot, but the design will have to be communicated with the primary users repeatedly. The primary users will give feedback about the design and the engineers designing the robot have to implement that feedback to optimize the design for the primary users. There has to be some kind of equilibrium when listening to the feedback. The feedback only consists of the interaction between the robot and the direct users. But the robot is an autonomous robot. The most important thing is that the robot can harvest the potatoes, and this should go without the humans touching it. That is, of course, besides the regular checkups which are discussed above under first users. When the robot gets designed this is all taken into consideration, but the secondary users have to take design decisions. These involve trade-offs between practicality for the autonomous operation of the robot and the human interaction.

#### Tertiary users

The tertiary users are the most important group that feels any effects of the robot. These are the investors, workers of, for instance, SpaceX and NASA and the director of the company designing the robot. First of all, we have the investors. Their main interest is to spread their name and maybe to be able to use the new technology later in, for instance, the crop cultivation sector here on earth. Like mentioned above, technology like this is expensive and there are a lot of different solutions. Most of these are cheaper and easier than a special robot. If there is a bigger reason, like a mars robot, than it actually is a good time to invest in an autonomous robot with an eye to the future. Investing in a project like this also gives your company a lot of attention, especially if the company provides certain parts for the construction of the robot. A project, like sending people to Mars fully equipped to make a civilization on the plains of that planet, will be a lot in the news, in magazines and more. People talk about certain technological advancements, that are made by a company, giving it a better name.

Companies, like SpaceX and NASA, can have multiple benefits if the robot becomes reality, but the most important is, of course, that when the robot exists, a lot of previous challenges are tackled. Keeping food for a long time in space without letting it spoil is done by using special astronaut food, however you do not want and you cannot use the same dried food again and again. One option is to do this by manual labor, but that is very time consuming and it is also heavy work in such a new environment. The robot would provide a fine solution that can be used eventually for multiple vegetables and fruits. The time that normally would be lost by manual labor is used to solve other challenges. The robot simply makes it much more accessible for the these kind of companies to do what was thought to be impossible.

The company that designs the robot is of course run by a director or a board. These people will not be in direct contact with the robot or the design process of the robot. But these people will feel the direct results of the completion of the robot. Like discussed earlier, the robot can be developed for different areas than only Mars. Only then the directors need to have some sort of plan on how to advertise for it. The directors should be in close contact with the designing engineers. In this way, the directors can communicate with the outside world on what the robot is capable of and on how to use it.

But before a director and its company can do these kind of things, he or she first has to make sure that it gets to work on this project. Because of logical reasons a company like NASA will not make the robot by itself. NASA would outsource a job like this to another company. All interested companies present their own plan on how to build it and their costs, etcetera. A director and a board will, of course, have a view on how their own company has to function over a long period of time. With that view they can decide if it is the right move to apply for the job. If the answer is yes, they can start by having contact with their engineers on how to realize the robot, what the total costs would be, and of course how to convince the so called judges to give the job to them. The directors and or managers are responsible for these kind of things and thus it is very important that they can be a good middle man in the development before, during and after the case.

#### User needs

In this section, the specific user needs of the above mentioned users are listed.

Primary users:

• UN1: Enough food available on Mars.
• UN2: In time, get a variety in food on Mars.
• UN3: Assembling the robot should be as easy as possible
• UN4: They should be able to do their jobs properly, without being too much involved in the autonomous greenhouse.
• UN5: They should have enough living and working space on Mars.
• UN6: They should be able to give feedback on the secondary users about the design decisions.

Secondary users:

• UN1: The design’s requirements should be made as clear as possible for the engineer.
• UN2: They should have enough time and money to come up with a good design for the robot.
• UN3: Their opinion about the robot’s design should be taken into account, as they have the most knowledge about it.

Tertiary users:

• UN1: Sponsors should be kept up-to-date about the process of designing the autonomous greenhouse and the robot.
• UN2: The companies get more brand awareness.
• UN3: The process of designing the robot should take as less time and money as possible.
• UN4: The technology used in this project can also be used as spin-off. (when looking into the future).

#### Scenarios

Steven, biologist, is already introduced in a prior scenario. He notices that it is a week since the last check has been executed in the greenhouse. It is time to take a look in the greenhouse whether all the potatoes are still doing well. He takes a fast glance through the greenhouse to take a look at some plants whether they look healthy. Furthermore, the robot has picked a random plant for investigation. Steven tests the plant for sicknesses and that the grow processes are going well and on schedule. When the plant is not healthy, Steven has to take a look at more plants to find out where the problem is. It could be a malfunction in irrigation or fertilization. It also could be the temperature regulator. Luckily there is nothing wrong with the plant and Steven goes on to his next task.

When Steven finds a plant contaminated with a dangerous pathogen, he immediately has to take samples of other plants to see how far this sickness has spread. If not all plants are contaminated, then these have to be sustained for new seeds and eventual consumption. The latter is only the case when enough plants survive. If the entire harvest is ruined, it has a great impact for the users on Mars. They now have to eat from an emergency food package which should be enough for at least 350 days, as this is approximately the time it takes to transport cargo to Mars (Drake. 2009). Earth is requested for a new package as the current one is in use, so the inhabitants on Mars have enough food if these kind of emergencies happen again. If all the plants die, the greenhouse needs to be sterilized for a new cycle. However this scenario is highly unlikely, as every single material is sterilized before it is brought to Mars. There ought to be a tiny chance that this happens. Therefore it is truly important that the primary users on Mars know how to act right, even for these kind of scenarios.

It is almost the end of the week and that means that the fertilization tank should be checked and refilled. This seems like a job for Tom, who is the ecologist. He has to fill the fertilization tank with the manures produced by the human digestive system. It is an easy job that is done in a couple of minutes. However, while checking the tank, he notices that the opening is obstructed by a dried blob of manures. Tom now has to clog the system to fix this. When this is done the tank is refilled, the robot is ready for the next round of fertilization.

The potatoes are full grown and ready for harvest. After the harvest, the potatoes are brought to a storage facility where they are stored for a period of time, but first, these potatoes need to be cured. This is also done in the storage facility. Steven has to check whether this is done right in the facility and that the potatoes are clean enough. As seed potatoes and the potatoes used for consumption need a different way of storage (as stated in the subsection “storage” above), a ratio of potatoes is separated for another storage room. Also this needs to be checked whether the right amount is separated. Further, as stated before, damaged potatoes have to be consumed first and should not be stored. Steven has to check in the storage after the curing process that the damaged and eventual diseased tubers are separated from the rest. If that is the case, the people on mars first eat the damaged potatoes and then get the potatoes out of storage every day for consumption.

When the robot in the greenhouse malfunctions, it cannot take care of the potatoes and they may die. It is therefore crucial that it functions all the time, because the people on Mars could have a severe lack of food if a harvest fails. It is therefore necessary that not only the robot contains a good feedback system for malfunctions, but that it is also checked regularly. This is done by the mechanical engineer Gary who is 39 years old. He checks whether all mechanical parts are still intact. Gary does this by performing the simple tests stated above in the subsection “primary users”. One of the tasks is pulling a full grown potato plant from the ground. This is done using a plastic model of a potato plant that is buried in the ground and then again dug up by the robot. When a part of the robot is broken, it is Gary’s job to fix it or replace it. If he would not be able to do that, he will consult the other mechanical engineers, like Harold, who might know what could fix the problem. If there is a software malfunction, it is Linda’s job to take a look and fix the problem, as she has a master in computational science.

During the phase of the initiation of tuberization the robot breaks down and it cannot immediately be repaired by the first users on Mars. This is a major inconvenience as now the inhabitants on Mars have to take care of the potatoes by hand which takes a great deal of time. Not only for the primary users this is inconvenient, but also for Matthew. The CEO of the company that was involved in creating the robot. He now has to evaluate how it is possible that the robot broke down. His company also loses stock value as the bad news is spread by the media. The investors are also unhappy as the market value of the shares are also lowering due to the bad news. Now the company needs to find an immediate solution to make the robot working again, as when that fails, the name of the company gets even worse.

### Model

Potato growth in Mars climate (ground, temp., rain and vapor pressure from Mars)
Potato growth in ideal climate (ground, temp. and vapor pressure from Earth)

For the estimation of the size of the robot, it is required to know the size of the greenhouse. To obtain this information, either tests or simulations have to be run. Testing the growth of the potato in an autonomous greenhouse on Mars is not yet an available option. Therefore it is better to run simulations.

Within this simulation model, one can change a number of variables. like explained in different parts of this page, the growth of the potato reacts to all of these in a different way. To model the numerous variables, some assumptions have to be made. However, this is because of the complexity of modeling a non-numerical process, or simply because of the lack of knowledge. Not all processes are examined and mapped out in a numerical way so it cannot always be modelled.

Starting with the growth of the plant itself, the growth will vary from plant to plant and this has to be taken in to account, since it will affect things like the water absorption. That absorption will affect the total needed water supply. Water absorption differs between the different phases of the plant growth, and at different times it will absorb more or less water. This is because in different phases the assimilation rate of the plant varies. A basic assumption is that when there is a higher assimilation rate, the needed water is also higher. The water retention and hydraulic conductivity of the soil itself are other variables that control the water management. This all creates a large loop, making up the water part of the model.

There are a lot more variables that are taken into account. Like discussed above, the properties of the ground are an important factor. The properties of how it interacts with water are important as well. The nutrient values are as well an aspect to take into account, affecting the model in a different way, since the regeneration value of the nutrition in the ground is also important. When the plants use all the nutrition, there has to be a way to refeed the ground so it matches the needs of the plant. Since the greenhouse on Mars will have a way to distribute the fertilizer, the refeed will be almost instant. This is necessary since the ground on Mars has no organic contents, so the Martian ground will not refeed the soil by itself.

The different variables used by the model are in data files, these files contain different sets of data, ranging over a certain time to keep somewhat realistic variable circumstances. The weather here on earth, of course, varies a lot, but in the greenhouse it can be controlled in a very precise way to achieve the best growth-time. With these optimal settings also the largest most nutritious potatoes are obtained.

The files of the potato model can be found here.

### Explanation results

#### Nutrition

As discussed in the section “manuring and nutrition” above, potatoes need certain nitrates that are in the Earths organic ground by nature. This is modelled with 3 different nutrients, these can be added with the fertilization equipment of the autonomous robot. The three nutrients are: Nitrogen (N), Phosphorus (P) and Potassium (K). These values can range from 0 up to 100 kg/ha and do not depend on each other. When experimenting with the different values a very easy pattern is to be found, the potatoes grow in a linear pattern parallel with the amount of nitrates that are in the ground as long as there is a balance between the different nitrates. So determining the optimal values for the nitrates in the ground is done easily, since their changes in values have independent consequences.

#### Martian ground

The variables concerning the ground are expressed in the model as ranges. The model uses the values for the water retention and hydraulic conductivity. The calculations are done in between the most extreme values that are found or calculated on Mars. This gives a first indication in what has to be added to increase the growth of the plants and what has to be done to increase the water flow.

#### Conclusion model

With knowing this all one can determine what is necessary to sustain enough potatoes needed for the people living on mars. Now certain decisions can be made to make an accurate representation of how it would be to sustain life on Mars

### Surface area of the potato field

The surface area (A) of the greenhouse is obtained from the simulation model. This model returns the mass of potatoes per hectare (σ). Then the required mass (M) of potatoes is needed to calculate the area with

$\displaystyle{ \mathrm{A = \frac{10000 \cdot M}{\sigma}}, }$

Where 10000 is a factor to convert the potato mass density (σ) from hectare to meters. Now the required mass is the value which should be obtained. This is simply done with

$\displaystyle{ \mathrm{M = \tau \cdot m \cdot p \cdot n}, }$

where τ is the amount of days the potatoes grow, which is also obtained from the simulation model. m Is the required mass of potatoes per person per meal, p is the number of human beings on Mars, and n is the amount of meals per day. n and p are assumptions. Now filling in the first formula by the second returns

$\displaystyle{ \mathrm{A = \frac{10000 \cdot \tau \cdot m \cdot p \cdot n}{\sigma}}. }$

This is only a batch of potatoes that is meant for consumption. There also should be grown some potatoes that are used as seeds for later batches. So from every plant, one potato is taken to be used as seed. To still fulfill the required mass for consumption, the current area (A) should be multiplied according to

$\displaystyle{ \mathrm{Afinal = \frac{10000 \cdot \tau \cdot m \cdot p \cdot n \cdot (1 + \frac{1}{k})}{\sigma}}. }$

Afinal is the final used area and k is the amount of tubers that averagely grows from one plant.

Now the final expression only has to be filled in using the following values;
m = 0.225 kg/(days*meals*people). Per meal men eat 0.250 kg of potatoes and women eat 0.200 kg (source: optima vita). The mean value of this is 0.225 kg.
p = 20 people, which is an assumption.
n = 3 meals, which is a standard amount of meals. It is also assumed that potatoes for now is the only available starch source on the planet.
k = 6.5 potatoes per plant, The amount of potatoes that are retrieved per plant is 3 to 10 (S. Behjati et al. 2013). The average of this is 6.5 assuming the distribution is poissonian.

These values result in a Surface area of 1650 m2.

This calculated surface area is the minimum amount of area needed. If something goes wrong with a couple of plants or tubers, then the users will already run too short. Therefore some margin is taken to make the size of the field 1760 m2. This is 5.5% of extra potatoes, which should cover for damage of other tubers and plants. With this surface area, the measures of the robot will be calculated so it is proportional. More about this will be explained in “scalability” and “design robot”.

With a surface area of 1760 m2, the measures of the field can be 44 meters long and 40 meters wide. Due to the fact that the width of the field is 40 meters, a robot which is 2 or 4 meters wide does not have to overlap tracks when plowing or harvesting(see "design robot"). The robot will be moving sideways at the end of the field, so there will be an additional piece of land where it is allowed to do that. Potatoes grow in ridges, so when moving sideways, the robot may not tramp them. For this an additional length of 6 meters is added to the greenhouse on both sides (In "design robot" will be explained why exactly 6 m). The total measures of the greenhouse then become 56 meters long by 40 meters wide.

### Design robot

The farming robot inside the greenhouse will be able to plant, harvest, and fertilize the potatoes. The body of the robot will be approximately 4 meters wide and 2 meters long. These dimensions are chosen because the bigger the robot gets, the more expensive it becomes. However, it still needs to move efficiently through the greenhouse and can therefore not be too small. A longer robot does not make it more efficient, wider does. But still some space is needed for the motor. Concluding to a robot between 2 and 7 meters wide. The robot would be most efficient if the field's width would be a multiple of the robot’s. Regarding the surface area of the field, the chosen width would be between 4 and 5 meters. Also, a width of 40 m and a length of 44 m would result in the right surface area. The width of the field is then a multiple of the robots as required. Further, 5 meters seems to be too wide for a total field width of 40 meters, so therefore a length of 4 meters has been chosen. In that way the robot only has to go back and forth ten times through the greenhouse without overlapping any of the ground. The robot must not be too high, since there is a sprinkler system at the top of the greenhouse. therefore, the maximum height is 3 meters. The farming robot might still be massive for the amount of field it has to cover, but that will be discussed later in "scalability".

The robot will look like a tractor, and it will thus do the planting, harvesting, and fertilizing in a similar way like common tractors do. The differences are that it will be able to do this autonomously and that both wheels will be the same size. For this, the robot needs a very good navigation. Otherwise, it will manage some parts of the field twice and other parts not at all. Therefore, it can track where it has already been and where it should go next. Which can be done by using a camera detection system (Bloch, 2017).

The robot will use tools that are attached additionally to the body. They will be attached to both sides and then the total length of the entire robot could be increased to 6 meters. As the robot needs to move sideways outside of the field to switch lanes, an additional piece of ground is needed, like discussed in "surface area of the potatoe field". This results in the total field measurements of 40 m by 56 m.

##### Fertilization

For the fertilization of the greenhouse, two different options are regarded. The first option was fertilization through a sprinkler system. This seemed efficient at first sight, since there already is a sprinkler system for the water supplies. But after this option has been inspected further, it turned out this might not be as efficient after all. This would have meant that there had to be a separate tank to connect the sprinkler system to, such that the system can switch between the two tanks. The idea here was to have two different pipes, merge in one pipe. In this case, if only water is needed, the fertilization pipe would close and vice versa. But this would bring a very high risk for leakage. This will result in a lot of maintenance, which is to be avoided as much as possible.

Therefore, a second option is chosen, which is implementing the fertilization into the robot. This is efficient, since the robot is already there. If the robot would only be used to plant and harvest potatoes, it would be a very expensive piece of machinery which will only be used about four times a year. Fertilization is a feature that has already been implemented in multiple tractors today. Using that feature, the robot will be of much more use. This option is also more scalable. If the greenhouse would be extended once the people on Mars get offspring, the robot can already cover an entire bigger greenhouse. Which means only the walls and the water sprinkler system need to be added.

##### Movement

Given the size of the greenhouse and the robot, it is important that the robot has a small turning radius. Therefore, the robot will be able to move forward, backward and sideward. For this it has two types of wheels. In the pictures below the robot is sketched with the different type of wheels. They are either defined as “type 1” or “type 2”. The robot has four wheels of each type.

The path of the robot is a straight line covering the entire length of the greenhouse. After that, the robot moves sideways to a whole new lane and it starts to overpass the length again. This will be repeated until the robot has covered the entire greenhouse. To overpass the length of the greenhouse, the robot uses its type 1 wheels. When the robot needs to move sideward, it will use the type 2 wheels. These can be moved in such way that they are positioned parallel to the front and the back of the tractor (Bac et al, 2014). In this case the type 1 wheels are lifted off the ground so that they do not interfere with the movement. When the robot reaches the end of the field, it lowers the type 2 wheels. After that it lifts the type 1 wheels in such way that it can move sideward to the next stroke of potatoes. Then the type 1 wheels and type 2 wheels do the same thing backwards and the robot can go further with its task.

There is however a downside. Since the robot does not rotate, the robot moves both for- and backwards. The planting and harvesting machinery works only in one direction. This might be seen as a big problem, but it is easily fixed with a second planting and harvesting tool. In this way the robot is basically mirrored in the lengths axis. The unused tool will be lifted if it is not being used. Leading to an advantage. Because the robot is mirrored, it is not useless when one of the two tools starts to fail. It just means the robot must travel twice as much. If such failure occurs, then the robot must move back the same lane it just covered to be able to cover the next lane. So then planting and harvesting starts from the same side every lane. That is an inefficient movement, since it overpasses every lane twice, but during that time the other side of the robot can be fixed.

##### Planting and harvesting

The arms of the robot will be able to plant and harvest potatoes. Planting proceeds in a similar way as the fertilization. For the planting a specific type of machinery is used, this device is called the Bomet Aardappelpootmachine [6]. The design will use a 4-meter-wide variant of the Bomet. When planting is necessary, the fertilization tool is detached from the robot and the Bomet is attached.

When the potatoes are ready, they can be harvested. This is much more complex than planting and fertilizing. In the video, [7], one of the most efficient ways to harvest potatoes is presented. The presented construction is not only too big, but it also needs trucks and truck drivers to make it really efficient. Therefore an exact replica is impossible to implement into the greenhouse. But the solution can be based on that video. To make it autonomous, humans should not have any direct influence on the process. Therefore the truck drivers, and also trucks, cannot be used. The harvesting machine, on the other hand, is a very useful piece of the puzzle. Currently the potatoes are transported by a treadmill and subsequently loaded into a passing truck. Since the greenhouse is not nearly as big as the field where the farmer in the video grows potatoes, there will not be the same amount of potatoes. Thus, instead of using the treadmill to load them intro a truck, the harvesting machine simply loads them into a big container around the machine. This is done in such a way, that the potatoes follow a similar route as in the video but the treadmill after the last corner is the container.

##### Maintenance

The primary users need to take care of the robot and make sure it is well maintained. The robot is constructed in such a way that the tools, which are used for harvesting, planting, and fertilizing, can be detached. When one of those parts needs maintenance, it can be extracted and easily fixed. The rest of the system can then still be up and running. In addition to the possibility of singling out the problem, the robot will get an extra room where it can be stored while inactive. Which is multifunctional since it can be used as a workshop for maintenance. Here back-up components are also stored, so the robot is complete always. In that workshop, the primary users can also contact engineers on earth about the robot.

The complete robot will be checked upon every month and the fertilization part will be checked upon every week. The robot should be cleaned and simple tests have to be done in order to find out whether the robot still functions well enough. For the cleaning, the users should ensure that the outlet of the fertilizer is empty and that the fertilizer tank is empty too. This will be done every single time after fertilizing. Once a month, the wheels of the robot will be checked upon by the users and the whole robot will be cleaned. In this way, the life time of the robot will be as optimized.

The simple tests mentioned before will be performed to ensure that the robot is still working. It will be mainly carried out by the programmer and the mechanical engineer, as they are specialized in this area. Examples of simple tests are given below:

• Pulling a full grown potato out of the ground to see whether the robot still does its job properly (done by biologists and agricultural experts)
• Let robot move forward and turn right and left
• Lift the wheels and lower them
• The opening of the tank including the fertilizer
• Checking the connection between the robot and the software (The working of this connection is also partly tested by performing the three above mentioned tasks)
• Checking whether the parts are well connected
• Checking the tire pressure

For cleaning the robot, water will be used, especially for the tank. This means that the materials used for the robot should be water resistant and the connections between the material parts have to be well sealed, such that no water will pass through.

#### Scalability

To have a robot, that only has to make 10 runs of 56 meters, to complete one whole circle is, of course, a bit overkill. The robot will be used for planting and harvesting the potatoes, which both have to be done once every 9 months. The only other task the robot has is to fertilize the crops once in a while, but this will also not keep the robot busy every moment of the day, on the contrary, the robot will not be operating most of the time. This is done with an purpose, because the Mars project is of course meant for the long run and not just to send one generation to Mars and abandon it afterwards. This is for all the different users not really ideal. Their interest lays also with the idea that life on Mars can be sustained and expended. A consequence of this is that the Mars population will grow.

To sustain the extra inhabitants on Mars a lot more food has to be produced to accompany for the increase in demand. At that moment the robot only provides the inhabitants with a very limited amount of potatoes. But since the robot is capable of much more, a few different options can be reviewed and considered to be the solution to the food shortage. The first option that can be considered, is expanding the greenhouse. With this idea, the total amount of potatoes can be increased. However, this causes the selection for the inhabitants of Mars to be stale and limited. When that is the case, the stay on Mars would not be so pleasant. But the robot can, of course, also be used for other ends than just potatoes. With the design that uses interchangeable attachments to make the robot do different tasks, new attachments can be designed to harvest other crops. Then the potato field will stay the same size, while a new greenhouse or an added part of the greenhouse will contain a new crop. This will make a more varied selection for the inhabitants of Mars. The job of the secondary users will also be easier. They only have to add new functions to the robot by sending new attachments and some new software to Mars. With some good communication to the inhabitants of Mars, the new attachments can be added easily without even needing real engineering or adjustments for the inhabitants of Mars. With the limited resources available on Mars, especially the first few years, nobody wants to do unnecessary complicated tasks. Heavy physical work also increases the danger level with minimal protection and resources. This work, of course, has to be avoided as much as possible.

An added bonus of increasing the variety of foods is the added nutrients that will be included in the more varied diet.

This is for the tertiary users also a very good option, since they manage and finance the project for a good part. It also is cheap for the tertiary users since the development costs can be kept to a minimum. This technology can also be used on earth as well, as is discussed in the section “Users”.

An added benefit is that life on Mars will become more as an Earth-like society. When people start seeing more development, it will become more attractive for them to go to Mars and live there. This can lead to a steep increase in the population on Mars, giving again, extra challenges to the secondary and tertiary users. This all makes it a very logical choice to make the robot a bit too capable for its job, since it can solve so much problems that otherwise would occur in the future.

#### Schematic drawings of the robot

Figure 1: 3D image of the robot including the fertilization tank and the front plough
Figure 2: Mechanism which shows the pulling up and lowering of the wheels
Figure 3: Mechanism which shows the folding in and out of the support rail
Figure 4: Front view of the robot not being attached to anything
Figure 5: Front view of the ferilization machine
Figure 6: Front view of the harvesting machine
Figure 7: Front view of the plough

Figure 1: 3D robot view:

On this schematic drawing, one is able to see the robot in 3D view. The plough has been attached to the robot together with the fertilization tank. It can be seen that the fertilizer tank sticks out a little bit behind the support rail. However, the fertilizer still stands firmly on the robot because the rest of the tank is well supported. Furthermore, the “click system” (which is the system that keeps the tools attached to the robot) ensures that the robot will not rotate on the support rails. On the drawing, it looks as if the fertilizer tank could fall over, however, the tires of the robot and the robot itself provide sufficient counterweight to ensure this will not happen.

Figure 2: Wheel lifting:

In this image, two drawings can be seen of the same scale: the front view and side view of a wheel, including the mechanism to lift them. The plank, which climbs up the rails, is attached to the cylinder holding the support around the wheel. What can be seen is that the rail along which the wheels will climb, is not that large. This is because it only has to be made sure that the wheel is entirely inside the robot, it does not have to be lifted more than necessary. Using the autonomous system, the robot knows when it has to move left, right or straight on. Knowing this, the robot automatically knows which wheels to pull up and which wheels to let down.

Figure 3: Side view retractable rail:

The block of the robot is shown in this figure, including the support rails. These rails can be folded out to support the harvest container and the fertilizer tank like a forklift truck. When the support rails are fold in, the plough on the back can be attached to the robot. The click system for this plough is located between the robot an the support rails when they are fold out.

Figure 4: Frontview robot only:

In this image, the wheels and click system on the front are shown. The plough can be attached to the robot by means of the click system. The inner and outer wheels have the same measurements. As can be seen, the wheels move in opposite direction. When the robot moves forwards, the inner wheels are pulled up and when the robot moves sideward the outer wheels are pulled up. The robot is drawn as a block here. All parts essential parts of the robot, like the motor, are located inside. The robot is quite large as shown in the figure. This space is needed for pulling up the wheels amongst other things.

Figure 5: Frontview fertilization tank:

The tank is supported in the same way as the harvesting machine is supported, by means of a rail and a click system. On this figure, the sprinklers are also shown, which are located beneath the tank. The pipes of the sprinklers are not made too long in order to let the fertilizer out at a higher level from the ground. In this way, the fertilizer will be distributed in a better way.

Figure 6: Frontview harvesting machine:

The pipes through which the potatoes are transported, is shown in this Figure. The click system including a rail is located in front of these pipes, to make sure that the tank sticks well to the robot. The shovels, which are shown, are located in the back of the machine. The tank here is used to gather the potatoes.

Figure 7: Plough frontview:

The ground level is also drawn on this figure, to show that the shovels are also moving into the ground. The shovels are a hollow pipe cone which is cut off at around 2/3.

#### Attaching Machinery

The click system shown in the images is based on magnetism. The simplified engineering idea is explained in the following reference; [8]. It uses multiple permanent magnets. When the poles are aligned in a specific way, the magnetic field collapses and it stops the clicking system from attracting any magnetic material. This is for when no tool is attached. Rotating a couple of magnets 180°, makes the whole system attract again. The clicking system will be designed in such way that it can rotate these magnets locally to turn the magnetic field on or off. With this mechanism, the robot can attach to or detach from different tools. This technique is very efficient since you do not need electricity or a big power source to attach the robot to tools. Repairs will barely be necessary since permanent magnets stay magnetic indefinitely.

All of the parts that need to be attached and detached are stored on platforms that are of the same height as the height of the robot on the place the equipment needs to be attached. This way, the robot can go to the first platform to deattach something and then move over to the next platform to attach the new equipment. For this there is no manpower needed.

### Path of the robot

Demonstration of the path model (video)

To demonstrate what the robot does and how it moves inside the greenhouse, a video of the model has been made. The model is made with NetLogo, because this software is very useful for making these kind of models. The model consists of a field, base station and the robot.

The user interface of the model consists of two buttons, two counters and one screen. The buttons are used for setting up the model and after that to run and pause it. The counters show how much crops are in the storage of the robot (‘Crops in robot’) or at the base (‘Crops in base’). On the screen, the position of the robot and the state of the patches of the field can be seen. The robot is defined by a tractor icon.

The robot drives a predefined route over the field to plant, fertilize or harvest the crops. This path is the most efficient path across the field, it is shown below. The red line is the path that robot drives each time. It can be seen that all patches are not even ploughed yet. Both counters are still at zero.

After one round, the whole field is ploughed and thus the field is ready for the next round of planting. The robot drives the same round over the field again and plants crops on all patches. After this round, two rounds of fertilization take place, so that the crops can grow. Then the crops get harvested in groups of ten. The robot has to go to the base when it is filled with crops of ten patches, because the storage is not big enough to carry all the potatoes. When the robot is full, it continues its path to an edge of the field. It then takes the shortest possible road back to the base and after that it starts a new round from the base again. When the last patch of crops is harvested, the robot finishes that round and starts ploughing again. This all is demonstrated in the figure at the right.

Patches of the field

The states of the field are divided into five states, road and a base station.
State zero is the starting state of the field, this is the state in which the crop is not yet planted and the soil is not ploughed yet.
State one is the state in which the soil just has been ploughed, but nothing has been planted yet.
State two is the state in which the crop is just planted.
State three is the state to simulate growing. Also the robot fertilizes the crops in this stage, which simply means that when the robot passes over a patch, it develops to the next stage.
State four is the final state, in this state the plant is fully grown and ready to be harvested. When the robot comes across a patch with this state, it harvests the crop and sets the patches state back to state zero. Then one crop is added to the value of ‘crops in robot’ till this value reaches 10 crops, then the robot is fully loaded. If the robot is fully loaded it does not harvest any crops anymore and just continues its route to the base station where it will drop of its load, the value of ‘crops in base’ then increases with 10 crops.
Five is road. This strip is not part of the field, because the robot is designed to plough, fertilize and harvest while driving forward or backwards. The robot also does not make turns, but moves sideways, as mentioned before. This way it cannot plough, fertilize or harvest this strip thus we made it a road.
Six is the base station at which the robot drops off its load, the storing of the potatoes happens here.

The NetLogo file can be found here and the software needed to run it can be found here.

### Conclusion

A concept for a robot for in an autonomous greenhouse, has been made. The robot is easy to assemble and disassemble. It works fully autonomously, only some simple tests must be done to ensure the robot still does its job properly. This is not at all time consuming, especially when everything works according to plan. The validation of the robot’s functionality is a short checklist that needs to be done by the users. Only when one or more aspects of the robot are flawed and it needs fixing, the validation take longer. However, every part of the robot has a duplicate since the back is the mirrored version of the front. This is efficient because the machinery only works in one direction and turning the entire robot is inefficient for both time and space. So, mirroring the front and back of the robot ensures it does not have to go over every patch twice. The different tools the robot uses are easily attachable and detachable using a magnetic ‘click’ system, where the alignment of a few simple magnets determine whether it attracts or does nothing at all. The ‘click’ system is very efficient and sustainable, since it does not require any electricity or power to keep the magnets turned on. Only the magnets, which never lose its magnetic strength, are used.

The greenhouse in total has been designed by making use of a simulation model for potato growth. The size of the greenhouse is determined based on the model, and therefore how much surface area is needed to provide food for every meal for every person on Mars. On the inside, the greenhouse has a regulatable temperature, UV light, minerals in the ground and water for the plants. In combination with the autonomous greenhouse, the autonomous robot, gives the users on Mars guaranteed food for every meal, with the least amount of effort.

Both the greenhouse and the robot are currently tuned to farm potatoes. However, in the future one can simply interchange the tools the robot uses. With these new tools it can take care of other type of crops. There is also the possibilty to adjust the few regulatable variables to fit the specific crop that is next up for farming. Looking ahead, the autonomous greenhouse and robot are able to provide a balanced meal for every inhabitant on Mars with barely any effort from the users.

### Discussion

The measurements of the robot have been chosen using the surface area of the field. This surface area is retrieved from a model that calculates the mass of potatoes that grow on a certain area. However, in this model, there was no dependency of potato breed. It is known that one breed can differ greatly from the other. This is stated multiple times in “environment potato”. In that part of the report, it is thus not the goal to indicate the exact measures of the surface area of the field, but it is the goal to demonstrate how these measures are determined. When the specific potato breed has been chosen, its properties can be modified in a similar simulation model. This will result in other values for the amount of days used for one growth cycle and also the mass of tubers retrieved in that amount of days. A different surface area is needed and possibly the measures of the robot should be modified.

Further, In the NetLogo model, the robot returns to a potato deposit unit. When returning, it takes the shortest possible route to get there. However, the model demonstrates that the robot will not directly go back to the place where it left, but it starts to run the track like it starts over entirely. This is not intended and the robot should take the shortest route back to the place where it stopped harvesting. With more available time, this should have been applied in the NetLogo model.

Also, now the robot is designed in such fashion that it does not rotate on the field, it moves backwards, sideways and forwards. This means that on both sides of the robot, a plough is attached when it is running. An alternative option would be that it is attached to only one plough, but the robot rotates 180 degrees when it reaches the side of the field. In this way the robot is 2 meters shorter and the length of the field can be 4 meters shorter because the robot needs less space to maneuver sideways. However it is not chosen to do so, as extra mechanical parts are included in the device. When that is done, there are more mechanisms that can wear and the robot should need some extra maintenance. Regarding the user aspect that the robot should be as simple as possible, it is chosen to refrain from using a rotating mechanism in the robot.

### Reference list

- "Als je ooit eens op mars zou kunnen wonen,zou je dan ook naar de aarde kunnen "bellen"" - Startpagina (2012). https://www.startpagina.nl/v/wetenschap/astronomie/vraag/391981/mars-wonenzou-aarde-bellen/

- Bac, C. W., van Henten, E. J., Hemming, J. and Edan, Y. (2014), Harvesting Robots for High-value Crops: State-of-the-art Review and Challenges Ahead. J. Field Robotics, [31: 888–911. doi:10.1002/rob.21525]

- B.G. Drake, “Human Exploration of Mars Design Reference Architecture 5.0” (2009) https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20090012109.pdf

- Giles, F. (2017). Agricultural robots no longer science fiction. Florida grower, 110(2), 12-13. Retrieved February 23, 2018, from [9]

- Hill, P. (2016). 9 robots designed to enhance the farm workforce. Farmers Weekly, 165(3), 66-70,72,75. Retrieved from [10]

- Victor Bloch, Avital Bechar, Amir Degani, (2017) "Development of an environment characterization methodology for optimal design of an agricultural robot", Industrial Robot: An International Journal, Vol. 44 Issue: 1, pp.94-103, [12] $\displaystyle{ Insert formula here }$