Mechanism Selection for Multi-Robot Task Allocation

Abstract

Abstract. The work presented here investigates how environmental features can be used to help select a task allocation mechanism from a portfolio in a multi-robot exploration scenario. In particular, we look at clusters of task locations and the positions of team members in relation to cluster centres. In a data-driven approach, we conduct experiments that use two different task allocation mechanisms on the same set of scenarios, providing comparative performance data. We then train a classifier on this data, giving us a method for choosing the best mechanism for a given scenario. We show that selecting a mechanism via this method, compared to using a single state-of-the-art mechanism only, can improve team performance in certain environments, according to our metrics.

A Study Assessing the Impact of Task Duration on Performance Metrics for Multi-Robot Teams

“…The work presented here extends our prior work in which we demonstrated the importance of mission execution and weighing various performance metrics when comparing task allocation mechanisms [5], [4]. Here we assess the impact of task duration.…”

“…Finding the optimal solution to an MRTA problem is known to be NP-hard [1], [2], so a popular family of strategies takes a marketor auctionbased approach. Auctions are useful because they can distribute workload amongst team members whilst reflecting preferences of individuals [2], [3], [4]. A local optimum is determined by each robot (labelled bidder) and these optima are collectively reconciled by an auctioneer.…”

“…Our prior work has focused on empirical analysis of auction mechanisms [5], comparing SSI, PSI and a baseline Round Robin (RR) (first come, first served) in experiments conducted on physical and simulated robots. We have defined a comprehensive set of performance metrics [10] and learned a model for selecting an appropriate mechanism given mission parameters [11].…”

Review on state-of-the-art dynamic task allocation strategies for multiple-robot systems

“…Market-based task allocation is a prominent multiple-robot task allocation strategy (Schneider et al, 2017). It imitates the market trading concept (Luo et al, 2015).…”

“…The auctioneer robot allocates the tasks to the least quoted robots. This strategy is implemented in both centralized and distributed robot's coordination network (Schneider et al, 2017). Market-based task allocation is broadly divided into singleitem auctioning and combinatorial auctioning.…”