Risk assessment and navigation of a wheeled mobile robot in nuclear applications

Abstract:

Abstract—Robots used for nuclear decommissioning often fail prematurely due to the extreme conditions encountered. Generally, the robots are deployed with a human-in-the-loop operator to intervene if problems occur. However, in some of the most critical situations, the response time of the operator is too slow to prevent an incident. In such cases, if the robot is capable of assessing the risk present with a given operation or path, it could take an appropriate action until an operator can assess the situation and decide on an action to take, or could
calculate and proceed with an alternative action. This raises two important research questions: how can the robot assess risks and how can the risks be incorporated into the robot’s architecture? This paper addresses these two issues. We propose a navigational framework based on risk-assessment where a wheeled-mobile robot is able to assess risks like radiation, temperature, battery power and collision. These risks are calculated at different levels of the framework depending on the nature of the risks. We use the Conditional Value at Risk (CVaR) metric for measuring the risk value. Preliminary results have shown that the risk-based robots have longer mission times compared to robots without
risk-assessment.

Download PDF