PRE2017 3 Groep11

From Control Systems Technology Group

(Difference between revisions)
Jump to: navigation, search
Line 151: Line 151:
The article proposes a solution to regulate the drone delivery fleets in urban area. In this solution they look at what they call logistics support. The drones a company uses in their delivery fleet are placed in different categories depending on their characteristics, like vertical land/take-off UAV’s. In these categories are different regulation rule that should be taken into account when designing the logistics support. They state that delivery drones are used mainly for one-day or same-hour delivery and that is where they put their focus. The logistic support is a balance between life cost, performance and operational availability. And it helps the system to behave in a satisfactory way for all stakeholders, e.g. company and costumers. The logistics support provides support technology, maintenance facilities and supply support. An logistics support used in the military: In this system the UAV’s decides to which base it will return based on its health level(high, medium, low). Different bases have different facilities for supply and maintenance and work to do.  Different Technical operation levels(TOL) are introduced to describe the task frequency and complexity, the amount of level depend on system. The type of drones, legislation, TOL, health level and environment of the applied drone fleet are used to develop an efficient system for the drones. How to decide on this system is described further in the article.  
The article proposes a solution to regulate the drone delivery fleets in urban area. In this solution they look at what they call logistics support. The drones a company uses in their delivery fleet are placed in different categories depending on their characteristics, like vertical land/take-off UAV’s. In these categories are different regulation rule that should be taken into account when designing the logistics support. They state that delivery drones are used mainly for one-day or same-hour delivery and that is where they put their focus. The logistic support is a balance between life cost, performance and operational availability. And it helps the system to behave in a satisfactory way for all stakeholders, e.g. company and costumers. The logistics support provides support technology, maintenance facilities and supply support. An logistics support used in the military: In this system the UAV’s decides to which base it will return based on its health level(high, medium, low). Different bases have different facilities for supply and maintenance and work to do.  Different Technical operation levels(TOL) are introduced to describe the task frequency and complexity, the amount of level depend on system. The type of drones, legislation, TOL, health level and environment of the applied drone fleet are used to develop an efficient system for the drones. How to decide on this system is described further in the article.  
===A cost-optimization model in multi-agent system routing for drone delivery===
===A cost-optimization model in multi-agent system routing for drone delivery===
 +
Kim, M., Matson, E.T. “A cost-optimization model in multi-agent system routing for drone delivery” 5115th International Conference on Practical Applications of Agents and Multi-Agent Systems, PAAMS 2017; Porto; Portugal; 21 June 2017 through 23 June 2017; Volume 722, 2017, Pages 40-51. https://www.scopus.com/record/display.uri?eid=2-s2.0-85021246581&origin=resultslist&sort=plf-f&src=s&st1=Drones++AND+deliver&nlo=&nlr=&nls=&sid=dfd90f7a89df9f7bf81cbae63504aeef&sot=b&sdt=b&sl=34&s=TITLE-ABS-KEY%28Drones++AND+deliver%29&relpos=20&citeCnt=0&searchTerm=
====summary====
====summary====
The paper focusses on the cost for the drone delivery system and does that concentrating at the total weighted distance traveled and the distance limits. Time is an important factor in optimizing the costs, but the demand for free shipping has bigger influence on the usefulness of drone systems. Thus the cost of a traveling drone is more important than time. To reduce these costs in the past A* search and heuristics where used to minimize the travel distance, and thus costs. The algorithm uses one place to pick up supply and get maintenance. The battery life is therefore important for the drones. The system looks to the possibility to place drones on busses(public transport)so they keep moving to their target while charging their battery. The algorithm looks for a way to merge bus routes and delivery routes, saving air space, time and costs. The experiment show promising results in uses bus routes and UAV’s together. It can reduce the total mileage up to 50%. (See the table and graphs)
The paper focusses on the cost for the drone delivery system and does that concentrating at the total weighted distance traveled and the distance limits. Time is an important factor in optimizing the costs, but the demand for free shipping has bigger influence on the usefulness of drone systems. Thus the cost of a traveling drone is more important than time. To reduce these costs in the past A* search and heuristics where used to minimize the travel distance, and thus costs. The algorithm uses one place to pick up supply and get maintenance. The battery life is therefore important for the drones. The system looks to the possibility to place drones on busses(public transport)so they keep moving to their target while charging their battery. The algorithm looks for a way to merge bus routes and delivery routes, saving air space, time and costs. The experiment show promising results in uses bus routes and UAV’s together. It can reduce the total mileage up to 50%. (See the table and graphs)

Revision as of 10:55, 21 February 2018

Coaching Questions Group 11


Wiki syntax
linebreak: < br >
list: *
sublist: **
numbered list: #

Contents

Problem statement and objectives

Problem statement

The usage of drones has become popular in various fields of work and also for recreational use. Drones have a promising perspective in a lot of different areas, like delivery and surveillance. For instance, delivery drones are already used by amazon today, but there are still a range of problems with implementing this technology. Multiple privacy and evasiveness concerns rise up with the use of these drones. Logistically speaking, handling drones and its packages in limited air space isn’t done in an efficient way adding cost to delivering. Security risk in hacking drones and drones crashing into objects and people. As well as the technical limitations of the drones itself, like the low battery life and the relatively low maximal weight of the packages. All these problems make using delivery drones less profitable and less desirable for companies and consumers.

Objectives

The group will look into a setup for delivery drones and its infrastructure in an urban area, taking into account all the problems described above. Possibly there are changes needed to the technology of the drone as well, this would be implemented into the setup. The setup will therefore aim to reach the next objectives:

  • Shorter delivery times
  • Find a way for robots to drop products in a safe but non-invasive way
  • When to use what type of drones (drone/car)
  • How to increase reliability
  • Package handling
  • Determine social impact
  • Satisfying all the stakeholders

Who are the users

  • Delivery services(companies)
  • Product sellers(companies)
  • Product buyers(consumers/companies)
  • Hospitals
    • emergency organ transportation
    • delivering life saving equipment
    • delivering Check up equipment/medication to patients and professionals

What do the users require

  • Fast and reliable service
  • Low prices
  • Privacy of products
  • non-invasive service

Approach, milestones and deliverables

Approach

First we will study the literature to figure out to what extent our case is already researched. After that we will attempt to determine the present day problems with the technology. Using the different theories, models and results from the literature study we try to solve all the problems described in the problem statement and deliver an setup for the infrastructure of an urban environment. Possibly tradeoffs have to be made between problems.

Milestones

  • Literature study
    • Sources read
    • Sources summarized
    • Coming up with solutions for:
      • Shorter delivery times
      • Find a way for robots to drop products in a safe but non-invasive way
      • When to use what type of drones (drone/car)
      • How to increase reliability
      • Package handling
      • Determine social impact
      • Satisfying all the stakeholders
    • Combining solutions in one setup
    • Finishing visual representation setup
    • Finishing visual representation new drone(optional)

Deliverables

  • Write a report
    • User analysis
    • Financial evaluation
    • Setup containing solution for all the objectives
    • Sketches/visual representation for the setup/new drone technology

Who's doing what

All together:

  • Write introduction
  • Update the wiki(weakly)

Everyone for themselves:

  • Search for sources (10-2)
  • Summarize sources (19-2)

Sources and summaries

Auto-Selection Of Package Delivery Location Based On Estimated Time Of Delivery

Toksoz, Tuna; Gulkaya, Tutku; and Price, Thomas, "Auto-Selection Of Package Delivery Location Based On Estimated Time Of Delivery", Technical Disclosure Commons, (December 22, 2016) http://www.tdcommons.org/dpubs_series/359

Summary

Package delivery systems restrict the delivery of packages under some conditions. Present technologies select a delivery location and time based on multiple factors:

  1. They use standard times and locations: 9:00-17:00 at work, 19:00-21:00 at home
  2. They look at historical data (success rates in certain windows and locations)
  3. They estimate the time it takes for the package to get ready when ordered.
  4. They look what delivery times are optimal, and choose the one with lowest cost.

Drone shipping versus truck delivery in a cross-docking system with multiple fleets and products

Tavana, M., Khalili-Damghani, K., Santos-Arteaga, F. J., & Zandi, M. H. (2017). Drone shipping versus truck delivery in a cross-docking system with multiple fleets and products. Expert Systems with Applications, 72, 93-107.
https://www.sciencedirect.com/science/article/pii/S095741741630687X

Summary

Robotic Aerial Vehicle Delivery System and Method

Sisko, M. (2015). U.S. Patent Application No. 14/565,348.
https://patents.google.com/patent/US20150158599A1/en

Summary

In order to facilitate safe and coordinated delivery by UAVs (unmanned aerial vehicles), this patent provides a landing platform. This platform is placed on a safe place, away from individuals, preferably above the roofline. The platform can sense dangerous situations, and communicate with the UAV to guide it, or abort the delivery if a safe landing is not possible.

Drone delivery models for healthcare

Scott, J., & Scott, C. (2017). Drone delivery models for healthcare.
https://aisel.aisnet.org/hicss-50/hc/global_health/7/

Summary

This paper focusses on delivery models for urgent healthcare packages, like blood, medicine, defibrillators, samples and vaccines. Prototype ambulance drones are already currently in use, and since drone technologies are getting better and better, this is a mere glimpse at what the future can do to speed up first-aid and healthcare as a whole. This paper provides 2 models for logistics networks, designed to provide timely delivery of healthcare. The model uses warehouses and drone nests. The packages are transported from the warehouses to the drone nests by land, and then brought to the final destination by drone.

Multi-Agent Path Finding with Payload Transfers and the Package-Exchange Robot-Routing Problem

Multi-Agent Path Finding with Payload Transfers and the Package-Exchange Robot-Routing Problem Hang Ma, Craig Tovey, Guni Sharon, T. K. Satish Kumar, Sven Koenig
https://dl.acm.org/citation.cfm?id=3016346

Summary

This paper looks into a way to minimize the time it takes for all packages within a set to be delivered, working with the following model of the world: each robot caries one package, 2 adjacent robots can exchange their packages and each package needs to be delivered at a certain destination.

Parcel delivery in an urban environment using unmanned aerial systems: a vision paper

Anbaroğlu, B., " Parcel delivery in an urban environment using unmanned aerial systems: a vision paper", ISPRS, (November, 2017) https://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/IV-4-W4/73/2017/isprs-annals-IV-4-W4-73-2017.pdf

Summary

The development of UAV’s started in the military. From there on, the technology spread to many other sectors that also benefit from it for performing various tasks. For this article, the use of drones for delivery purposes is considered. For now, the public still has a rather negative image of drones, as most people fear privacy violations. Professional UAV pilots, however, more fear the risk of possible accidents. The use of drones to deliver goods also has a lot of benefits. Firstly, they replace trucks, which saves a lot of space on the road and pollution. Thereby, drone delivery is good for the economy and employment, because of all development and manufacturing needed to make drone delivery systems. In short, it is cheaper, faster, safer and better for the environment. In order to get to a real implementation of drone delivery, there are three important challenges to overcome: public acceptance, regulations and technology. For public acceptance, the biggest challenges are privacy, safety and reliability. Considering regulations, the biggest problems are the tracking of the UAV’s, the identification of non-authorised UAV’s, hacking prevention and routing of the UAV’s, in order to prevent accidents. The technological challenges are mostly related to the battery life, avoiding other things and big buildings blocking GPS signals.

The Sky’s (Not) the Limit - Influence of Expertise and Privacy Disposition on the Use of Multicopters

Lidynia, Chantal; Philipsen, Ralf; Ziefle, Martina, " The Sky’s (Not) the Limit - Influence of Expertise and Privacy Disposition on the Use of Multicopters ", Springer, (June 21, 2017) https://link.springer.com/chapter/10.1007/978-3-319-60384-1_26

Summary

The development of UAV’s started in the military. From there on, the technology spread to many other sectors that also benefit from it for performing various tasks. For this article, the attitude of certain groups of people towards drones is considered. In general, the attitude towards drones is rather positive. There are some concerns, however. The general public is mostly concerned about privacy, whereas professional drone pilots are mostly concerned about possible accidents. It is also mentioned that ‘privacy’ is a rather vague, ill-defined, thing, that also changes as a result of new technologies and other developments. Looking at the current legal situation, it becomes clear that none of the current legal restrictions for private drone usage is perceived as overly restrictive. Some general concerns towards drones(apart from the ones already mentioned) are the ue of drones as a weapon, misuse by criminals and spying. One last thing that is discussed by the article is the fact that people only really agree by drones flying over their houses is in case of rescue operations. For some other things, such as drone delivery, people only slightly approve this.


Post-Production Analysis Approach for drone delivery fleet

Asma, Troudi; Addouche, Sid-Ali; Dellagiy, Sofiene; El Mhamedi, Abderrahman; "Post-Production Analysis Approach for drone delivery fleet", IEEE Explore digital library, (December 01, 2017) http://ieeexplore.ieee.org/document/8120986/

Summary

Nowadays, a lot of companies are investing in drone delivery services, such as DHL with PaketKopter, Amazon with Amazon PrimeAir , Google with Project Wing and recently GeoPoste with Geodrone. In order to get drone delivery working good, there needs to be a very good management and Logistics Support in order to be reliable, cheap and fast enough. The subject of this article is the Logistics Support part of this. Logistics support should consists of two parts according to this article:

  • Prerequisite SL recommended by the manufacturer through an operating manual. (This is not enough, so this paper proposes the following:)
  • Post-Production logistics support system which depends on the exploitation requirements

The article also states that ‘’the identification of the fleet status, the health status or also maintenance facilities, helps to support the system during the exploitation phase’’. Also, maintenance is important in order to get an as high as possible availability. Logistics Support Analysis is used to increase efficiency of maintenance and reduces the cost of providing support by preplanning all aspects of Logistics Support. In this analysis, the article also suggests a new part, in order to also be able to take future changes into account (For example, when a drone fleet is equipped with a new battery, there will be a lot of implications for the entire infrastructure). This new part is called an Activity Analysis. In this Activity Analysis, all implications of a possible future change are evaluated and prepared for. Lastly, a model is made to optimize the total travelling distance of the UAV’s.

A multi-objective green UAV routing problem

Coelho, Bruno N.; Coelho, Vitor N.; Coelho, Igor M.; Ochi, Luiz S.  ; Haghnazar K., Roozbeh; Zuidema, Demetrius;. Lima, Milton S.F; da Costa, Adilson R., "A multi-objective green UAV routing problem", Elsevier, (December, 2017) https://www.sciencedirect.com/science/article/pii/S0305054817301028?via%3Dihub

Summary

The development of UAV’s started in the military. From there on, the technology spread to many other sectors that also benefit from it for performing various tasks. Nowadays, the biggest problems for the development of delivery drones are the very strict rules for drones (no flying in urban environments, drone must stay in sight of pilot, etc.) and the limited battery life of a drone. This paper considers a Green UAV Routing Problem (GUAVRP), which minimizes seven objective functions: total traveled distance; UAV’s maximum speed; number of used vehicles; makespans of the last collected and delivered package; average time spent with each package; and maximize batteries load at the end of the schedule. For the model, it is assumed that there are two flight levels: one low level for smaller drones and a higher level for bigger ones. In the map, there are also several refueling stations. According to the article ‘’The main contributions of this current work are:

  • Develop a mathematical programming model for a time- dependent UAV heterogeneous fleet routing problem, in particular:
    • respecting UAVs operational requirements;
    • tackling the micro-airspace considering a multi-layer scenario with package exchanging points;
    • integrating UAVs into the new concepts of mini/microgrid systems, in which vehicles can be charged at different points of the future smart cities.
  • Consider a multi-objective optimization framework in order to provide alternative solutions with different possible routes and schedules.’’

There are no real conclusions in this paper, it is mostly about making the model itself.

The Vehicle Routing Problem with Drones: Extended Models and Connections

Poikonen, Stefan; Wang, Xingyin; Golden, Bruce, “The Vehicle Routing Problem with Drones: Extended Models and Connections”, Wiley online library, (June 7, 2017) http://onlinelibrary.wiley.com/doi/10.1002/net.21746/full

summary

This article considers a truck that is carrying drones, so it is a combination of both drone and truck delivery. In short, they looked at the difference between a lot of slow drones and some fast drones. They found that if there are times when not all drones are in service (service not fully parallelized), greater drone speed dominates and if drone range or capacity is severely limited, a larger number of drones may dominate. Also, they stated that a drone will, for any destination, always travel a shorter (or just as long) path than a truck, because a drone isn’t bound to roads. Then, they also considered some routing problems, namely Close-enough-vehicle-routing-problem(CEVRP), vehicle-routing-problem-with-drones(VRPD) and vehicle-routing-problem(VRP) (CEVRP and VRP being known problems, and VRPD being the problem of this article). They found that CEVRP gives the optimal solutions of VRPD when the velocity of the drones is infinite and that VRP gives the optimal solution of the VRPD when the velocity of the drones is 0.

Logistics support for a delivery drone fleet

Troudi, A., Addouche, S.-A., Dellagi, S., El Mhademi, A. “Logistics support for a delivery drone fleet”, 2nd International Conference on Smart Cities, Smart-CT 2017; Malaga; Spain; 14 June 2017 through 16 June 2017, Volume 10268 LNCS, 2017, Pages 86-96. https://www.scopus.com/record/display.uri?eid=2-s2.0-85020887649&origin=resultslist&sort=plf-f&src=s&st1=Drones++AND+deliver&st2=&sid=dfd90f7a89df9f7bf81cbae63504aeef&sot=b&sdt=b&sl=34&s=TITLE-ABS-KEY%28Drones++AND+deliver%29&relpos=18&citeCnt=0&searchTerm=

summary

The article proposes a solution to regulate the drone delivery fleets in urban area. In this solution they look at what they call logistics support. The drones a company uses in their delivery fleet are placed in different categories depending on their characteristics, like vertical land/take-off UAV’s. In these categories are different regulation rule that should be taken into account when designing the logistics support. They state that delivery drones are used mainly for one-day or same-hour delivery and that is where they put their focus. The logistic support is a balance between life cost, performance and operational availability. And it helps the system to behave in a satisfactory way for all stakeholders, e.g. company and costumers. The logistics support provides support technology, maintenance facilities and supply support. An logistics support used in the military: In this system the UAV’s decides to which base it will return based on its health level(high, medium, low). Different bases have different facilities for supply and maintenance and work to do. Different Technical operation levels(TOL) are introduced to describe the task frequency and complexity, the amount of level depend on system. The type of drones, legislation, TOL, health level and environment of the applied drone fleet are used to develop an efficient system for the drones. How to decide on this system is described further in the article.

A cost-optimization model in multi-agent system routing for drone delivery

Kim, M., Matson, E.T. “A cost-optimization model in multi-agent system routing for drone delivery” 5115th International Conference on Practical Applications of Agents and Multi-Agent Systems, PAAMS 2017; Porto; Portugal; 21 June 2017 through 23 June 2017; Volume 722, 2017, Pages 40-51. https://www.scopus.com/record/display.uri?eid=2-s2.0-85021246581&origin=resultslist&sort=plf-f&src=s&st1=Drones++AND+deliver&nlo=&nlr=&nls=&sid=dfd90f7a89df9f7bf81cbae63504aeef&sot=b&sdt=b&sl=34&s=TITLE-ABS-KEY%28Drones++AND+deliver%29&relpos=20&citeCnt=0&searchTerm=

summary

The paper focusses on the cost for the drone delivery system and does that concentrating at the total weighted distance traveled and the distance limits. Time is an important factor in optimizing the costs, but the demand for free shipping has bigger influence on the usefulness of drone systems. Thus the cost of a traveling drone is more important than time. To reduce these costs in the past A* search and heuristics where used to minimize the travel distance, and thus costs. The algorithm uses one place to pick up supply and get maintenance. The battery life is therefore important for the drones. The system looks to the possibility to place drones on busses(public transport)so they keep moving to their target while charging their battery. The algorithm looks for a way to merge bus routes and delivery routes, saving air space, time and costs. The experiment show promising results in uses bus routes and UAV’s together. It can reduce the total mileage up to 50%. (See the table and graphs)

Understanding security threats in consumer drones through the lens of discovery quadcopter family

summary

The paper addresses privacy, security and safety issues of using drones, and the paper recommend basic steps to solve these issues. The article focusses on the drone technology and how to take over drones. This isn’t useful for our project. What is important for our project is the following paragraph. The article states the general concerns that people have with consumers drones, but also delivery drones. These are the privacy that the drones can fly over your piece of land, make videos/photo’s without you even seeing the drone. There are security concerns, the drones can stop work and fall down, not everyone is capable of correctly handling a drone. Most people won’ t misuse the drones, however the article states that most drone types are easily hackable and give multiple examples how. Sometimes only Wi-Fi connection and using the app, easily obtainable from the appstore of google play, can be used to take over the drone. The hackers get complete access to the movememt and camera of the drone. And can use them to take pictures, and crash it into places they want. Drones can also get to places where the operator wouldn’t be able to see before. Therefore new restriction for drones need to be thought of. The idea was to restrict the drone from getting to certain GPS locations, but with hacks this was easily avoided. The key conclusions of the article is basically that the security of consumer drones need to be improved significantly. More specific conclusions for the system are listed here:

  • Securing drone access point with a strong password, and WPA2.
  • Limiting the number of devices allowed to connect to the accesspoint. Also, enforcing user authentication, and denying income and outgoing traffic from and to unauthorized devices.
  • Disabling ftp and telnet.We found that none of these services are needed for the normal operation of the Discovery drones. But if there must be an anonymous ftp user, then the device should not allow read and write access to the entire root directory.
  • Sending network packets between app and drone over a secure channel.
  • Upgrading the software running in the device. The Discovery quadcopters (released in 2016) use BusyBox 1.20.2 which was released in 2012. Since then, there has been 18 software updates to BusyBox, and these devices may be vulnerable to other known BusyBox vulnerabilities [4]. (Unfortunately, it is not possible for users to update the firmware in Discovery drones).

Drone-Aided Healthcare Services for Patients with Chronic diseases in Rural Areas

Kim, S.J., Lim, G.J., Cho, J., Côte, M.J. “Drone-Aided Healthcare Services for Patients with Chronic diseases in Rural Areas” Journal of Intelligent and Robotic Systems: Theory and Applications Volume 88, Issue 1, 1 October 2017, Pages 163-180. https://www.scopus.com/record/display.uri?eid=2-s2.0-85017116386&origin=resultslist&sort=plf-f&src=s&st1=Drones++AND+deliver&st2=&sid=dfd90f7a89df9f7bf81cbae63504aeef&sot=b&sdt=b&sl=34&s=TITLE-ABS-KEY%28Drones++AND+deliver%29&relpos=5&citeCnt=1&searchTerm=

summary

Rural areas generally have less facilities then urban areas. For people with chronic diseases in rural areas it can be time consuming and costly to travel to hospitals every day or week for treatment and check-ups. The same can be said for the professionals that travel from home to home to give adequate home treatment. Drones give a good perspective for a solution on this problem, since the drones can deliver checkup kits, medication and more to hospitals and homes. Saving a lot of time and money for hospitals and patients. Furthermore, the professionals have more time for treatments were they actually have to be there, increasing the satisfaction of patient and the professionals themselves. The article contributes these points to the already existing literature:

  • Proposes a concept for drone-aided healthcare delivery and pick-up system in rural area
  • Model with two steps
    • Strategic planning to optimally decide drone centers
    • Operational planning for optimal drone flight
  • Tool to analyse cost-benefit
  • Provide an algorithm to solve proposed models

For the model and cost-benefit analysis could be useful to use in our own setup. However, the useful parts of the model would be the results and less the algorithm that the writer made.

Personal tools