A Distributed Version of the Hungarian Method for Multi-Robot Assignment
For multi-robot systems, this work provides a distributed algorithm for optimal assignment, addressing scalability and communication constraints in peer-to-peer networks.
The paper proposes a distributed version of the Hungarian Method for multi-robot assignment, enabling robots to cooperatively compute optimal assignments via local computations and communications. The method is applied to multi-robot routing with spatio-temporal constraints, demonstrated through an interactive orchestral floor experiment.
In this paper, we propose a distributed version of the Hungarian Method to solve the well known assignment problem. In the context of multi-robot applications, all robots cooperatively compute a common assignment that optimizes a given global criterion (e.g. the total distance traveled) within a finite set of local computations and communications over a peer-to-peer network. As a motivating application, we consider a class of multi-robot routing problems with "spatio-temporal" constraints, i.e. spatial targets that require servicing at particular time instants. As a means of demonstrating the theory developed in this paper, the robots cooperatively find online, suboptimal routes by applying an iterative version of the proposed algorithm, in a distributed and dynamic setting. As a concrete experimental test-bed, we provide an interactive "multi-robot orchestral" framework in which a team of robots cooperatively plays a piece of music on a so-called orchestral floor.