Papers
12
Total Citations
1,396
H-Index
7
About
Keith Sullivan is a computer scientist whose research spans multi-agent simulation, swarm robotics, autonomous systems, and machine learning. He is perhaps best known as a principal contributor to **MASON** (Multi-Agent Simulation Of Neighborhoods), a fast, extensible, discrete-event multi-agent simulation toolkit written in Java. Originally introduced in 2004 and refined in a landmark 2005 paper that has since garnered over 1,000 citations, MASON has become a foundational platform for researchers modeling everything from swarm robotics to social complexity systems, reflecting its broad and lasting influence on the field. Beyond simulation infrastructure, Sullivan has made meaningful contributions to robot autonomy and learning. His work on learning from demonstration enables robots and multi-agent teams to acquire complex collective behaviors through supervised examples, while his anomaly detection research—leveraging generative adversarial networks and deep neural networks—advances the capabilities of autonomous patrol robots. He has also explored practical human-robot collaboration, including touch-based communication for Navy firefighting teams and real-time learning for robot soccer. Sullivan's portfolio reflects a consistent drive to bridge theoretical multi-agent modeling with real-world robotic applications, making his work valuable to students and researchers across artificial intelligence, robotics, and autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1MASON: A Multiagent Simulation Environment1,007 citations · 2005
- 2MASON: A New Multi-Agent Simulation Toolkit258 citations · 2004
- 3Finding Anomalies with Generative Adversarial Networks for a Patrolbot45 citations · 2017
- 4Collaborative foraging using beacons20 citations · 2010
- 5Learning from demonstration with swarm hierarchies16 citations · 2012
- 6
- 7Detecting Anomalous Objects on Mobile Platforms12 citations · 2016
- 8Real-Time Training of Team Soccer Behaviors7 citations · 2013
- 9
- 10Representing motion information from event-based cameras4 citations · 2017