Technical design requirements of a safety reasoning module for run-time risk reduction in Highly Automated Off-road Mobile Machinery

• Automation shifts operators’ roles to event-based intervention supervisors. • Cabin presence adds safety burdens, needing greater risk reduction for automation. • SRM feasibility depends on hardware, software, and Machine Safety standards. • Insights guide SRM development despite challenges in specific automation cases. Highly automated off-road mobile machinery requires reliable safety systems to mitigate work- place risks, particularly when operating in close collaboration with human actors. With the ad- vent of automation, the operator’s role has shifted to that of a supervisor, intervening only during event-based scenarios. This supervisor may be situated either in the cabin or remotely, overseeing multiple machines. This article explores the technical design requirements and feasibility of a safety reasoning module (SRM) through a targeted use case involving the automation of a cut-to-length tree harvester. Specifically, it examines the automation of the log-feeding process required for stem inspection. The SRM’s primary function is to achieve adequate risk reduction while preserving the supervisor’s capacity for effective event-based intervention, thereby supporting a compelling safety case for maintaining a safe operational state at runtime. Currently, the supervisor plays a crucial role in the stem inspection process, requiring cabin presence which adds to the burden placed on required minimum risk reduction and event-based intervention. In the context of this evaluation the realisation of the SRM’s requirements for achieving the necessary risk reduction while conforming to Machine Safety standards indicates it is constrained by limitations in existing hardware and software capabilities. Despite these limitations, this use case outlines a structured approach for developing SRMs for highly automated off-road machin- ery. While achieving the desired risk reduction may not be feasible in this specific scenario, the findings provide valuable insights that can guide the development of SRMs in other applications where similar challenges arise.