Defensa de tesi de Daniel Rivas Alonso

Doctorand: Daniel Rivas Alonso.

Títol: Execution-driven Dynamic Multi-Robot Task Allocation Model easily Applicable to Real Cases.

Directors: Lluís Ribas Xirgo

Data i hora lectura: 08/02/2023, 11:00h.

Lloc lectura: Sala de Graus de l’Escola d’Enginyeria – Edifici Q.

Programa de Doctorat: Enginyeria Electrònica i de Telecomunicació.

Departament on està inscrita la tesi: Departament de Microelectrònica i Sistemes Electrònics.

 

Abstract

Today indoor logistics rely on fleets of autonomous mobile robots that transport goods from pick-up spots to drop-off areas. The number and variability of transport operations makes it impractical to manually allocate these tasks to robots, and their continuous flow and the unpredictability of the traffic conditions increase complexity of exhaustive search optimization task allocation techniques and make their results suboptimal if applicable. Practical solutions use auction-based protocols which, though may not give optimum, minimum cost allocations, adapt very well to dynamic scenarios. In this work, a parameterized auction-based multi-agent system is elaborated so to improve results of a basic auction mechanism for task allocation. The experimental results with physically simulated plants, including one with a typical warehouse-like layout and another from a SEAT automaker’s workshop [https://youtu.be/gOqv4b7YmlY], show that the total transportation cost of a set of time-ordered tasks in a given scenario can be reduced when adjusting auction parameters and mode of operation. Particularly, when auctions are repeated to include new unoccupied robots into the bidders’ set.