Simultaneous Configuration Formation and Information Collection By Modular Robotic Systems
Advisor Information
Prithviraj Dasgupta
Location
UNO Criss Library, Room 232
Presentation Type
Oral Presentation
Start Date
4-3-2016 9:45 AM
End Date
4-3-2016 10:00 AM
Abstract
We study a central problem in modular self-reconfigurable robots - the configuration formation problem - given a set of modules initially distributed arbitrarily within the environment and a desired target configuration involving those modules, how can each module select an appropriate spot or location in the target configuration to move to, so that, after reaching the position, it can readily connect with adjacent modules and form the shape of the desired target configuration. We study the additional navigation criterion for information collection while forming configurations - the modules have to select their navigation paths so that they can increase the amount of information they collect using their on-board sensors, while they are moving towards their positions in the target configuration. Information that the modules can collect from the environment can be of different types such as temperature measurement, algae sample collection, rock/soil sample collection etc. depending on the available on-board sensors of the modules. Each module has a limited energy budget to expend while moving from its initial to goal location. To solve this problem, we propose a budget-limited, heuristic search-based algorithm that finds a path that maximizes the entropy of the expected information along the path. We have analytically proved that our proposed approach converges within finite time. Our experimental results show that our planning approach has lower run-time and fewer messages exchanged than an auction-based allocation algorithm for selecting modules’ spots.
Simultaneous Configuration Formation and Information Collection By Modular Robotic Systems
UNO Criss Library, Room 232
We study a central problem in modular self-reconfigurable robots - the configuration formation problem - given a set of modules initially distributed arbitrarily within the environment and a desired target configuration involving those modules, how can each module select an appropriate spot or location in the target configuration to move to, so that, after reaching the position, it can readily connect with adjacent modules and form the shape of the desired target configuration. We study the additional navigation criterion for information collection while forming configurations - the modules have to select their navigation paths so that they can increase the amount of information they collect using their on-board sensors, while they are moving towards their positions in the target configuration. Information that the modules can collect from the environment can be of different types such as temperature measurement, algae sample collection, rock/soil sample collection etc. depending on the available on-board sensors of the modules. Each module has a limited energy budget to expend while moving from its initial to goal location. To solve this problem, we propose a budget-limited, heuristic search-based algorithm that finds a path that maximizes the entropy of the expected information along the path. We have analytically proved that our proposed approach converges within finite time. Our experimental results show that our planning approach has lower run-time and fewer messages exchanged than an auction-based allocation algorithm for selecting modules’ spots.