# Difference between revisions of "Feedback to user needs, requirements and preferences"

(Created page with 'The performance indicators in the last three columns of the results of Table 1 are important values for the primary and secondary users, mentioned in Chapter 3. The requirements …') |
|||

Line 23: | Line 23: | ||

earnings on water-savings. If the individual client is the supplier of the water, which is likely, the | earnings on water-savings. If the individual client is the supplier of the water, which is likely, the | ||

chance that the window cleaning company could change its algorithm for a particular job based | chance that the window cleaning company could change its algorithm for a particular job based | ||

on feedback could increase. | on feedback could increase.<br/><br/> | ||

The last important requirement for the algorithms is to stay below an energy consumption of 73.5 | |||

''W''. From the results the only algorithm which comes closes to this value is the Standard algorithm | |||

with a value of 70.6 ''W'', which would mean that all the algorithms meet the requirement. The | |||

most energy-efficient algorithm is the turndirt algorithm, followed by the zigzag algorithm. The | |||

manufacturer would be able to use the same batteries as for the state-of-the-art window cleaning | |||

robots. The same goes for the water tank.<br/><br/> | |||

Altogether, the results indicate that the zigzag algorithm is the the fastest cleaning algorithm, | |||

the turndirt algorithm the most water- and energy efficient algorithm. Both algorithms beat the | |||

reference algorithm on all fields. This means that the manufacturer has a choice in whether to | |||

choose for either the most time-efficient or the most energy- and water-efficient algorithm. |

## Latest revision as of 20:07, 2 April 2018

The performance indicators in the last three columns of the results of Table 1 are important values
for the primary and secondary users, mentioned in Chapter 3. The requirements of no water
stripes and minimal coverage, are implemented in the algorithm script beforehand and are
achieved this way. However, the performance requirements have still to be met. A cleaning
speed of a speed of 125 *m2/hr* was desired. From the results however, it can be seen that this
goal is far from reached. With a maximum of 27.30 *m2/hr*, the zigzag algorithm is the closest
to the goal. This large difference in cleaning speed can be explained through the interpretation
of the definition of cleaning speed. It is assumed that the manufacturers meaning of the term is
the surface area covered by the robot, but not specifically fully cleaned. The current modeled
algorithms cover specific areas multiple times in order to clean the window. This repeatedly traveled
distance is not taken into account in the definition of the cleaning speed as derived from the
results of the model. Besides that, a contradiction emerges when the cleaning speed and the
reference algorithm are compared. The standard algorithm represents the state-of-the-art algorithm,
however it does not hold up with the specifications of the state-of-the-art robots. Despite
not corresponding with the state-of-the-art specifications, the zigzag algorithm is still faster than
the reference algorithm, meaning that the time taken to clean a window is shortened which is
beneficial for both window cleaning companies and for individual clients.

The requirement for the water consumption was equal to 0.25 *L/min*. From the results it can be
seen that both developed algorithms are far from reaching this value, which is a positive result.
For the window cleaning companies, the choice to either save time or water is now based on the
results of the test. However, it is assumed that the profit on time-savings is more relevant than the
earnings on water-savings. If the individual client is the supplier of the water, which is likely, the
chance that the window cleaning company could change its algorithm for a particular job based
on feedback could increase.

The last important requirement for the algorithms is to stay below an energy consumption of 73.5
*W*. From the results the only algorithm which comes closes to this value is the Standard algorithm
with a value of 70.6 *W*, which would mean that all the algorithms meet the requirement. The
most energy-efficient algorithm is the turndirt algorithm, followed by the zigzag algorithm. The
manufacturer would be able to use the same batteries as for the state-of-the-art window cleaning
robots. The same goes for the water tank.

Altogether, the results indicate that the zigzag algorithm is the the fastest cleaning algorithm,
the turndirt algorithm the most water- and energy efficient algorithm. Both algorithms beat the
reference algorithm on all fields. This means that the manufacturer has a choice in whether to
choose for either the most time-efficient or the most energy- and water-efficient algorithm.