Luis Enrique Pineda

Evolving communication in robotic swarms using on-line, on-board, distributed evolutionary algorithms

Robotic swarms offer flexibility, robustness, and scalability. For successful operation they need appropriate communication strategies that should be dynamically adaptable to possibly changing environmental requirements. In this paper we try to achieve this through evolving communication on-the-fly. As a test case we use a scenario where robots need to cooperate to gather energy and the necessity to cooperate is scalable. We implement an evolutionary algorithm that works during the actual operation of the robots (on-line), where evolutionary operators are performed by the robots themselves (on-board) and robots exchange genomes with other robots for reproduction (distributed). We perform experiments with different cooperation pressures and observe that communication strategies can be successfully adapted to the particular demands of the environment.

Equi-Reward Utility Maximizing Design in Stochastic Environments.

We present the Equi-Reward Utility Maximizing Design (ERUMD) problem for redesigning stochastic environments to maximize agent performance. ER-UMD fits well contemporary applications that require offline design of environments where robots and humans act and cooperate. To find an optimal modification sequence we present two novel solution techniques: a compilation that embeds design into a planning problem, allowing use of off-the-shelf solvers to find a solution, and a heuristic search in the modifications space, for which we present an admissible heuristic. Evaluation shows the feasibility of the approach using standard benchmarks from the probabilistic planning competition and a benchmark we created for a vacuum cleaning robot setting.

Continual planning for search and rescue robots

The deployment of robots for emergency response tasks such as search and rescue is a promising application of robotics with growing importance. Given the perilous nature of these tasks, autonomous robot operation is highly desirable in order to reduce the risk imposed on the human rescue team. While much work has been done on creating robotic systems that can be deployed for search and rescue, limited work has been devoted to devise efficient real-time automated planning algorithms for these tasks. In this work, we present REDHI, an efficient algorithm for solving probabilistic models of complex problems such as search and rescue. We evaluate our algorithm on the search and rescue problem using both an abstract domain representation and a semi-realistic simulator of an existing robot system. The results show that REDHI can obtain near optimal performance with negligible planning time.