James Su, Sebastian Schweigert, Sunil Srinivasan, Jiewen Sun and Mark Jouppi
EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2015-89
May 14, 2015
http://www2.eecs.berkeley.edu/Pubs/TechRpts/2015/EECS-2015-89.pdf
To truly make automation economical, there needs to be a shift in robotic performance: from highly specific at one task to general-purpose across many tasks. Unfortunately, such additional functionality is not cost-effective without a sacrifice in performance. In the case of robotics, it is a sacrifice in accuracy. This paper first investigates an industry analysis of a general-purpose robot to outline its merits and drawbacks as a competitive product. The paper then delves into more technical detail about how to accommodate the robot's lack in accuracy. Specifically, our method of implementation is to have the robot self-detect when it is too inaccurate to proceed with a task and subsequently request assistance from the human. Here, we utilize the fine-grained capabilities of a human, relatively, in synergistic combination with the efficiency of an inexpensive robot.
Advisor: Ruzena Bajcsy
BibTeX citation:
@mastersthesis{Su:EECS-2015-89, Author = {Su, James and Schweigert, Sebastian and Srinivasan, Sunil and Sun, Jiewen and Jouppi, Mark}, Title = {Robotic Manipulation with a Human in the Loop – Accuracy Comparison and Analysis}, School = {EECS Department, University of California, Berkeley}, Year = {2015}, Month = {May}, URL = {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2015/EECS-2015-89.html}, Number = {UCB/EECS-2015-89}, Abstract = {To truly make automation economical, there needs to be a shift in robotic performance: from highly specific at one task to general-purpose across many tasks. Unfortunately, such additional functionality is not cost-effective without a sacrifice in performance. In the case of robotics, it is a sacrifice in accuracy. This paper first investigates an industry analysis of a general-purpose robot to outline its merits and drawbacks as a competitive product. The paper then delves into more technical detail about how to accommodate the robot's lack in accuracy. Specifically, our method of implementation is to have the robot self-detect when it is too inaccurate to proceed with a task and subsequently request assistance from the human. Here, we utilize the fine-grained capabilities of a human, relatively, in synergistic combination with the efficiency of an inexpensive robot.} }
EndNote citation:
%0 Thesis %A Su, James %A Schweigert, Sebastian %A Srinivasan, Sunil %A Sun, Jiewen %A Jouppi, Mark %T Robotic Manipulation with a Human in the Loop – Accuracy Comparison and Analysis %I EECS Department, University of California, Berkeley %D 2015 %8 May 14 %@ UCB/EECS-2015-89 %U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2015/EECS-2015-89.html %F Su:EECS-2015-89