ML safety assurance scoping

Objectives

1. Define the scope of the safety assurance process for the ML component.
2. Define the scope of the safety case for the ML component.
3. Create the top‐level safety assurance claim and specify the relevant contextual information for the ML safety argument.

Inputs to the stage

• [A] : System safety requirements
• [B] : Description of operating environment of system
• [C] : System description
• [D] : ML component description
• [F] : ML assurance scoping argument pattern

Outputs of the stage

• [E] : Safety requirements allocated to ML component
• [G] : ML safety assurance scoping argument

Description of the stage

As shown in the AMLAS ML assurance scoping process diagram above, this stage consists of two activities that are performed to define the safety assurance scope for an ML component. The artefacts generated from this stage are used to instantiate the ML safety assurance scoping argument pattern as part of Activity 2. An ML component comprises an ML model (e.g. a neural network) that is deployed onto the intended computing platform (i.e. comprising both hardware and software).

Additional guidance on the use of ML for autonomous systems can be found at [9].