Background

Recent advances in robotics and computer science mean that it is now possible to use teams of robots for many real-world applications. One very important application is robotic search-and-rescue. Robots are highly useful for search and rescue tasks since they may be deployed in dangerous environments without putting human responders at risk. Moreover, robots can explore locations that are hazardous or inaccessible to humans (such as mine fields, contaminated zones, or underwater environments). While much effort is going into development of more robust and mobile robot platforms, it is also necessary to develop practical user interfaces for humans to control such a team of robots. Efficient coordination of a multi-robot team requires advanced methods for communication and cooperation.
The Oxford University Computing Laboratory is currently cooperating with the Intelligent Systen Laboratory Amsterdam to develop a simulated multi-robot search-and-rescue team. This team integrates advances from a variety of fields such as navigation, mapping, exploration, human-robot interface design, communication and cooperation.
All simulation work is performed using the USARSim framework. Based on a commercial game engine, this simulator uses state of the art techniques for simulating physics and rendering graphics. It is fully configurable, allowing for independent development of simulated robot models, sensors and environments.

Assignment

As the robots explore an environment they create a map of what they have sensed. These maps are used to coordinate the exploration process, either autonomously or by a human operator. Precise maps are very important and prevent robots from colliding with obstacles (or victims). Currently the robots use a laser range scanner to create a 2-dimensional map. Image feedback from a camera can be used by human operators to detect and avoid obstacles.
It would be of great interest to develop a 3-dimensional mapping system. 3-Dimensional laser range scanners are not yet common but it is possible to use a 2-D laser scanner for this purpose with the addition of a mount and a servo-motor. The main goals of this project would be to develop a simulated model of such a modified 2-D laser range scanner, to validate its performance in simulation, and if time permits to use the acquired data for creation of 3D maps.

Tasks

• Read relevant publications about the USARSim simulation environment
• Develop an understanding and awareness of current 3D mapping techniques
• Install the USARSim simulator and the most recent Amsterdam-Oxford code release
• Learn how to manipulate the simulator
• Build a simulated model of a laser range scanner on a mount with a servo
• Test the model extensively
• Use the acquired data from the simulated 3D scanner to estimate the traversibility of the environment

Recommended Readings

• S. Carpin, M. Lewis, J. Wang, S. Balakirsky, C. Scrapper (2007). USARSim: a robot simulator for research and education. Proceedings of the 2007 IEEE Conference on Robotics and Automation. PDF
• S. Thrun. Robotic mapping: A survey. In G. Lakemeyer and B. Nebel, editors, Exploring Artificial Intelligence in the New Millenium. Morgan Kaufmann, 2002. PDF
• H. Surmann, A. Nuchter, and J. Hertzberg, An autonomous mobile robot with a 3D laser range finder for 3D exploration and digitalization of indoor environments, Robotics and Autonomous Systems, vol. 45, pp. 181-198, 2003. PDF
• David M. Cole and Paul M. Newman Using Laser Range Data for 3D SLAM in Outdoor Environments. International Conference on Robotics and Automation, 2006 Florida PDF
• A. Visser, T. Schmits, Steven Roebert and Julian de Hoog, "Amsterdam Oxford Joint Rescue Forces - Team Description Paper - Virtual Robot competition - Rescue Simulation League - RoboCup 2008", Proceedings CD of the 12th RoboCup Symposium, Suzhou, China, July 2008. PDF