Sensor Networks and Adaptive Reconfigurable Systems (SNARL)

Laboratory


Adaptation and Learning in Multi Robot Teams

Agency: DoD Army Research Office (2003 - 2006)
This research focusses on the hardware and software architectures for the realization of teams of intelligent mobile Autonomous Ground Vehicles. The architecture should allow for the modular and hierarchical approach to implement deliberative and reactive behaviors in teams of autonomous vehicles. A hiearchical architecture called Adaptation and Learning at All Levels is developed that enables:
* intelligent sensor selection and sensor fusion,
* fault detection and accommodation,
* dynamic reconfigurability and high levels of reliability under all operating conditions,
* hierarchical architecture for implementing hybrid controllers,
* sophisticated algorithms for implementing adaptive behaviours,
* remote operation and coordination with other controllers, and
* provision for system evolution, and paths for the migration of capabilities.

Sensors for the Inspection and Monitoring of Gas Pipelines

The overall objective of the proposed research is to develop cost-effective technology for inspection and monitoring of gas pipelines using sensor networks. The proposed development will focus on energy requirements, optimization, locomotion, and implementation of networks for collecting data by sensors for inspection and monitoring of gas pipelines. The developed system will be versatile in order to accommodate for various applications and it will provide a practical and economic means of collecting and transferring various types of useful measured data. This not only facilitates inspection and integrity management of gas pipelines traveling over long distances in difficult to reach areas, but also reduces the cost of compliance with different federal and state regulations.