Carnegie Mellon Robotics Institute
The broad topic that enfolds my research could be called repairable or reconfigurable robots. I am interested in robots that can repair themselves, reconfigure themselves, repair a teammate, reconfigure a teammate or build other identical robots. Any technology that would be required for these capabilities is of interest to me.
In my earlier work I looked at building some simple robots that could perform a limited form of repair. The design was intentionally kept very simple. The idea here was to demonstrate that repair is not a difficult thing to achieve and to gain some insight on what design features make for a good repairable robot.
I am currently working on determining the expected gain in utility of having a team of repairable robots versus having a team of non-repairable robots. This theoretical analysis with several unrealisitic assumptions is simply to determine an upper bound on how much more work you can expect a team of repairable robots to do versus a team of non-repairable robots made with the same components. This work is not yet published as it is still in progress.
Please refer to my personal web page.
|Research Interest Keywords|
|game theory, machine learning, multi-agent systems, planning|
|The Robotics Institute is part of the School of Computer Science, Carnegie Mellon University.|
Contact Us | Update Instructions