Our New TUBITAK 1001 Research Project

Project Title : Effective task allocation for autonomous heterogeneous multi-robot search and rescue team

Use of robots for search and rescue missions in case of disasters is of great importance in terms of security of the rescue personnel, and quick response to victims. Academic studies are pursued around the world for more effective and secure rescue robots.
Scope of the presented project covers development and use of autonomous heterogeneous multi-robot team that can detect victims and report back about their well-being, in disaster environments. Heterogeneous multi-robot team is consists of a central ground vehicle, aerial vehicles and other ground vehicles and is planned to be implemented for indoor/interior areas. Each member of the multi-robot team would be equipped with diverse set of sensors and mobility abilities. So high coordination among heterogeneous robots is expected to result in more flexible and cost efficient infrastructure compared to high sensor ability and mobility capable homogeneous multi-robot teams.
Search and rescue with heterogeneous multi-robot team requires solutions for various sub-problems. Among these sub-problems effective task allocation is at most importance with respect to requirement for coordinated utilization of partial abilities such as local mapping, local exploration, local victim detection and local obstacle avoidance. As an original approach task based exploration is going to be implemented for the solution of coordinated exploration for multi-robot team with presented project. Moreover to be able to speed up navigation path planning in 3D space, navigation path cache and utilization of 3D JSP and D* algorithms is proposed to be used as original implementations. Use of depth data as a source of visual odometry in 3D mapping is also expected to work better compared to RGB visual odometry in case of insufficient lighting.
Outcomes of the project such as developed methods and infrastructures are going to be tested in international competitions and that would hopefully lead to contributions to academic community and better recognition of Turkey in field of rescue robotics.

Keywords: heterogeneous multi-robot team, search and rescue, task allocation, task oriented coordinated exploration