Project Group/Seminar/Lab Report: Putting Things On A Table Efficiently
Project Group/Seminar/Lab Report: Putting Things On A Table Efficiently
Abstract 3 Methodology
Every day we pick things up from the ground and put them We will now discuss our novel approach in detail.
on tables. Despite the ubiquitous presence of this task, this
problem does not have a sufficiently optimized solution yet. 3.1 Robot
In this report, we will explore the usage of robotics in order
to make putting things on tables more efficient. The base of the robot is a cube with wheels. In order to be
able to grab things, it also has a robotic arm. At the end of
1 Introduction the arm is a robotic hand with three fingers.
The work of First and Second (2014) and others (Third, The algorithm is very simple and consists of the following
2015) is fundamental in this research area. They show how steps:
difficult it can be to pick things up and put them on a table.
All of these approaches work, but none of them is as 1. Grab something using the pickup() function
efficient as ours, as can be seen in Section 4.
1
(a) The robots body. (b) The robots arm.
Data sets avg. speed avg. error which are novices and some of which are experts in putting
things on tables. Table 1 shows the average pickup speed of
Robot 125.1 2.2 our test subjects compared to the pickup speed of our robot.
Human 87.4 2.1 During our experiments we discovered, that people being
supported by our robot are happier than people, who have to
Table 1: Pickup speed and error comparison. grab their stuff themselves. This is qualitatively evaluated in
Figure 3.
5 Future Work
While our robot is already able to outperform most of our
test subjects in a fully autonomous manner, it still can be
improved with respect to the mobility. Right now, due to the
Tears small wheels, it is not suited in situation with a potentially
very bumpy ground, e.g. gardening.
Robotic Support
6 Conclusion
We have developed a novel robotic solution to the problem
No Tears of picking things up and putting them on a table. Despite
the shortcomings described in Section 5, we still believe
in widespread usage of the presented robot within the next
decade.
4 Evaluation
In order to find out, how useful the proposed robot is in
reality, I will do a user study with 500 participants, some of