BITE: Bar Ilan Teamwork Engine
Since 2004, The MAVERICK group has been investigating general team decision-making mechanisms for multi-robot teams, pursuing joint goals. The driving hypothesis underlying this work is that while the specific protocols and procedures for reaching joint decisions may vary from one task to another (who does what? when do we start? when do we stop?), there are general invariants which allow automation of the decision-making process.
Put simply, we are interested in how to make robots behave well in a team, regardless of what task the team is to carry out. BITE is software which implements the insights from these investigations: It has evolved over the years, from one thesis to the next.
BITE pioneered significant changes in the control loop that is the backbone of agents’ decision-making, allowing them to switch social choice protocols on the fly. It was the first teamwork control software to provide proven guarantees on the coherence of the team. In software engineering studies, it was shown to save up to 60% of software development efforts. Ideas from BITE publications have influenced commercial uses in projects by Cogniteam and Intuition Robotics.
For more information about BITE and general teamwork for robots, see the following publications:
- G. A. Kaminka and I. Frenkel. Flexible teamwork in behavior-based robots. In Proceedings of the Twentieth National Conference on Artificial Intelligence (AAAI-05), 2005.
- G. A. Kaminka and I. Frenkel. Integration of coordination mechanisms in the BITE multi-robot architecture. In Proceedings of IEEE International Conference on Robotics and Automation (ICRA-07), 2007.
- G. A. Kaminka, A. Yakir, D. Erusalimchik, and N. Cohen-Nov. Towards collaborative task and team maintenance. In Proceedings of the Sixth International Joint Conference on Autonomous Agents and Multi-Agent Systems (AAMAS-07), 2007.
- G. A. Kaminka. No robot is an island, no team an archipelago: Plan execution for cooperative multi-robot teams. In ICAPS 2015 Workshop on Planning and Robotics (PlanRob), 2015.