University of Manchester
Browse
1/1
5 files

Avoidance of Single Gamma Source in Simulation

dataset
posted on 2022-01-21, 15:19 authored by Andrew WestAndrew West
Simulation results from a Jackal UGV commanded to navigate over a point gamma radiation source. The UGV and gamma source were in an open simulated environment, therefore able to take any diversion the path planner generated.

The robot has three possible states of knowledge of the radiation field: no information, known knowledge, or unknown knowledge. These three states are labelled as "no_avoidance", "known" and "unknown" respectively.

In the no avoidance scenario, the robot is not made aware of any radiation information. In the known environment, the robot is driven manually around the environment to build up a map of the radiation field, before being asked to perform navigation tasks. In the unknown case, the robot has no prior knowledge, but can react to new radiation information as it completes a navigation task.

By adjusting the upper thresholds of cost associated with radiation risk, the robot is more or less cautious (there is lower or higher cost associated with navigating through regions of increased radiation dose rate).

When the robot is aware of the radiation field, its accummulated dose is reduced, at the expense of greater time required to complete the task. Reduction in dose is readily an order of magnitude compared to the no_avoidance case.

Data is in comma separated format, with header. Header format is: time (s), x (m), y(m), z (m), value (arb units). Value represents radiation intensity.

Funding

TORONE - TOtal characterisation for Remote Observation in Nuclear Environments

Engineering and Physical Sciences Research Council

Find out more...

Robotics and Artificial Intelligence for Nuclear (RAIN)

Department for Business, Energy and Industrial Strategy

Find out more...

History