Discipline: Technology and Engineering
Subcategory: Computer Science & Information Systems
Joseph Gonzalez - UCI
The effectiveness of search and rescue robotics in large scale areas is currently limited by the battery-lifetime of autonomous, mobile units since they must return to the start location to have their batteries manually replaced; this is time wasted not looking for disaster victims. This problem is difficult to solve since large areas require more time to be mapped and mobile robots do not yet possess a means of recharging themselves in a varying environment. We seek to amend the problem by incorporating light-finding behavior into a mobile unit’s environment mapping process for later pairing with solar panels on the unit. The behavior will be refined in a simplified, indoor environment with a defined energy harvesting area and global fixes, in the form of QR codes, for correcting location estimations in the environment. The goal of this research is to teach a mobile unit how to pick an energy harvesting area based on light-finding behavior. A desired area has the characteristics of having direct access to a source of sunlight and closeness to the area currently being mapped. Implications for results include increased efficiency at exploring larger areas with lessened need for human intervention.
Not SubmittedFunder Acknowledgement(s): LSAMP
Faculty Advisor: Jeff Krichmar, jkrichma@uci.edu
Role: I read up on the Kalman filter, wrote the code, tested it on the robot, ran the experiment, and wrote up the paper.