市場ベースのマルチロボットタスク割当を救急車配車に応用
The Application of Market-based Multi-Robot Task Allocation to Ambulance Dispatch
市場ベースのマルチロボットタスク割当手法を救急車配車問題に適用し、ロンドン救急サービスの100万件超の履歴データで評価したところ、現行の手動配車より応答時間が統計的に有意に改善した。
著者: Eric Schneider, Marcus Poulton, Archie Drake, Leanne Smith, George Roussos, Simon Parsons, Elizabeth I Sklar
分類: cs.MA, cs.RO
原文アブストラクト
Multi-Robot Task Allocation (MRTA) is the problem of distributing a set of tasks to a team of robots with the objective of optimising some criteria, such as minimising the amount of time or energy spent to complete all the tasks or maximising the efficiency of the team's joint activity. The exploration of MRTA methods is typically restricted to laboratory and field experimentation. There are few existing real-world models in which teams of autonomous mobile robots are deployed "in the wild", e.g., in industrial settings. In the work presented here, a market-based MRTA approach is applied to the problem of ambulance dispatch, where ambulances are allocated in respond to patients' calls for help. Ambulances and robots are limited (and perhaps scarce), specialised mobile resources; incidents and tasks represent time-sensitive, specific, potentially unlimited, precisely-located demands for the services which the resources provide. Historical data from the London Ambulance Service describing a set of more than 1 million (anonymised) incidents are used as the basis for evaluating the predicted performance of the market-based approach versus the current, largely manual, method of allocating ambulances to incidents. Experimental results show statistically significant improvement in response times when using the market-based approach.
