A Decentralised Strategy for Heterogeneous AUV Missions via Goal Distribution and Temporal Planning

Yaniel Carreno, Èric Pairet, Yvan Petillot, Ronald P. A. Petrick

PosterID: 1 PDF Slides BibTeX

Heterogeneous robot platforms offer the potential to support complex missions, such as those needed for persistent autonomy in underwater domains, since the robots can be optimised for specific tasks in order to better manage dynamic contexts. While current temporal AI planners are able to deal with multi-agent planning problems, by producing plans that take into account the individual robot capabilities and task requirements, these approaches often manage the high-dimensionality of the state space in such problems inefficiently, leading to multi-agent plans with poor plan quality. This paper proposes a novel decentralised task allocation strategy called Decentralised Heterogeneous Robot Task Allocator (DHRTA) which enables the computation of less complex plans for individual robots using temporal planning. The strategy decomposes a set of mission goals by considering their spatial distribution, execution time, and a robot's capabilities to reduce the computational cost resulting from considering large numbers of possible goal assignments. Experiments illustrate the robustness of the approach and indicate improvements in plan quality by reducing the planning time and mission time, while significantly reducing the rate of mission failures.

Session E5: Robotics & Embedded Applications
Canb 10/27/2020, 17:00 – 18:00
10/31/2020, 00:00 – 01:00
Paris 10/27/2020, 07:00 – 08:00
10/30/2020, 14:00 – 15:00
NYC 10/27/2020, 02:00 – 03:00
10/30/2020, 09:00 – 10:00
LA 10/26/2020, 23:00 – 00:00
10/30/2020, 06:00 – 07:00