Repurposing a Roomba: Evaluating and training behavior in a simple agent

Recent attempts to reprogram a Roomba to be used as a simple agent have led to interesting behavior. Observation has shown that the behavior of the Roomba is not only dependent on the precepts of the Roomba, but also relies heavily on the uncontrollable environmental conditions that the Roomba is placed in. Ultimately this makes the Roomba a great platform to test and teach aspects of artificial intelligence. This paper will show how most of the tested environmental conditions are mitigated by a learning agent that will adjust behavior dependent on the precepts that are received.

Name thesis:degree_name

Master of Science

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Computer Science

Year dc:date.available

2007

Author dc:creator

• Abbott-McCune, Donald Samuel

Contributors dc:contributor

• Dr. David Primeaux

• repurposing
• simple agent
• Roomba
• artificial intelligence
• Computer Sciences
• Physical Sciences and Mathematics

Statement dc:rights

• © The Author

Identifier

https://scholarscompass.vcu.edu/etd/1077

OAI identifier oai:identifier

oai:scholarscompass.vcu.edu:etd-2076