RAIN Robot operation SC Strawman

Generated from ASCE 5.0.8 on 09/03/2022 at 15:27

Author: Chris Anderson

Version: 1.0.0

Description: This Robotics with Autonomy/AI Safety Case has beend generated as an example framework, based on a hypothetical robot operating in a nuclear material storage pond. Therefore, the safety case contains placeholder nodes. e.g. the Evidence nodes would normally contain links/references to documents. This safety case provides two examples of how the 'Collision' hazard may be mitigated (Method 1 and Method 2). It is unlikely that both methods would be needed to show that the risk has been reduced to be ALARP and, therefore, one could be deleted. In the event that both are needed to show defence in depth and diversity then Method 1 and 2 would need to be restructured into one set of CAE. It is widely recognised that the use of autonomy/AI in robotics in high integrity applications can be difficult to justify and therefore create a robust safety case. This safety case is based on the premise that, as a minimum, tasks and hazards are identified/analysed early in the robot's design/development lifecycle, thereby reducing the likelihood that decissions will be made which would preclude a robust justification. This safety case is an ettempt to begin a discussion between the Robotics/Autonomy community (many in academic research), industry/commerce and the nuclear regulator in the UK (Office for Nuclear Regulation (ONR)). Therefore, it may only form part of the solution, however, it is reasonably certain that Method 1 can be used now and it is feasible that Method 2 can be used in the medium term, providing the necessary verification tools and methodologies are developed. This Safety Case was generated using an ASCE ACADEMIC LICENSE and therefore cannot be used for commercial applications


User views for this network:

1. Top View
2. Method 1
3. Method 2