A Closed-Loop Bid Adjustment Method of Dynamic Task Allocation of Robots

Abstract

Task allocation in dynamic environments is one of the most challenging problems in multi-robot coordination, and it is imperative for a wide range of applications. Auction-based approaches are popular methods that aim to assemble robot team information at a single location to achieve practicable task allocation. However, a main deficiency of auction-based methods is that robots generally do not have sufficient information to estimate accurate bids to perform tasks, particularly in dynamic environments with operational uncertainties. While some techniques have been developed to improve bidding, they are mostly open-looped without feedback adjustments to tune the bid prices for subsequent tasks of the same type. Robots’ bids, if not adjusted accordingly, may not be trustworthy and would indeed impede team performance. To tackle this issue, we propose a closed-loop bid adjustment mechanism to improve robots’ bids, and hence enhance the overall team performance. Each robot in a team maintains its own track record as closed-loop feedback information to adjust its bid prices. After a robot has completed a task, it assesses its performance to reflect the discrepancy between the bid price and the actual cost of the task. A series of such performance records, with time-discounting factors, are taken into account to damp out fluctuations of bid adjustments. Adopting this adjustment mechanism, a task would be more likely allocated to a competent robot that submits a more accurate bid price, and hence improve the overall team performance. Simulation of task allocation of automated guided vehicles serving at a container terminal is presented to demonstrate the effectiveness of the bid adjustment mechanism.