This paper describes the steps taken to create the score criteria aimed at measuring the quality of maps produced by teams participating in the RoboCup Rescue Virtual Robots competition. Since metrics have already been developed by a few research groups, we start by highlighting the most popular solutions to this problem, emphasizing their strengths and weaknesses. Having put the difficulty of creating map benchmarks into perspective, we present our map benchmark suite, appropriate for Urban Search and Rescue missions, along with examples taken from former competitions.
Proceedings Title: Proceedings of the Performance Metrics for Intelligent Systems (PerMIS) 2009
Conference Dates: September 21-23, 2009
Conference Location: Gaithersburg, MD
Pub Type: Conferences
Map evaluation, RoboCup, Performance Metrics, Performance, Measurement