The University of Pennsylvania MAGIC 2010 multi-robot unmanned vehicle system

Abstract

In this report, we describe the technical approach and algorithms that have been used by the University of Pennsylvania in the MAGIC 2010 competition. We have constructed and deployed a multi‐vehicle robot team, consisting of intelligent sensor and disrupter unmanned ground vehicles that can survey, map, recognize, and respond to threats in a dynamic urban environment with minimal human guidance. The custom hardware systems consist of robust and complementary sensors, integrated electronics, computation, and highly capable propulsion and actuation. The mapping, navigation, and planning software is organized hierarchically, allowing autonomous decisions to be made by the robots while enabling human operators to interact with the robot team in an efficient and strategic manner. The ground control station integrates information coming from the robots as well as metadata feeds to focus the attention of the operators and respond rapidly to emerging threats. These systems were developed and tested by the UPenn team to complete two phases of the MAGIC 2010 challenge in a safe and timely manner.