Multi-robot coordination in a warehouse

From ISLAB/CAISR
Title Multi-robot coordination in a warehouse
Summary The project aims to coordinate motion of multiple robots to accomplish a single task
Keywords multi-robot, path planning, spatial conflict
TimeFrame
References [[References::[1] Andreasson H, Bouguerra A, Cirillo M, Dimitrov DN, Driankov D, Karlsson L, Lilienthal AJ, Pecora F, Saarinen JP, Sherikov A, Stoyanov T. Autonomous transport vehicles: where we are and what is missing. IEEE Robotics & Automation Magazine. 2015 Mar;22(1):64-75.

[2] Gombolay MC, Wilcox R, Shah JA. Fast Scheduling of Multi-Robot Teams with Temporospatial Constraints. InRobotics: Science and Systems 2013 Jun 24.]]

Prerequisites Robotics, Intelligent Vehicles, Algorithm courses
Author
Supervisor Jennifer David, Björn Åstrand
Level Master
Status Open


In this project, we aim to resolve the spatial conflicts that arise when multiple robots work together in a warehouse scenario. This involves a higher level task assignment system that provides conflict-free paths for each of the robot to accomplish a single task such as scanning the warehouse together or moving pallets to different places [1]. The problem of task assignment is similar to vehicle routing problem with time-window constraints and is NP-hard. They are usually solved using genetic programming or other heuristic approaches. When this problem is coupled with multi-robot coordination, it has become a very important research question these days due to the automation of different warehouses all over the world (Ocada, Amazon KIVA, GoCart, etc.). In this project, we try to prune the many solutions to the NP hard problem by taking into account the knowledge of the environment (spatial) as additional constraints. This work is directly an extension of the paper [2].The TERCIO method developed at CSAIL, MIT is the starting point of this project, where we will work on adapting this algorithm to work with spatial conflicts and more dynamic environment such as a warehouse.