The main focus of the SUCCESS project was the assurance of safety for a quarry site case where numerous machines interacted and collaborated. The team conducted research on hazard/safety analysis methods, adapted them to the quarry site requirements, and incorporated best practices into a new assurance methodology that is also applicable to similar classes of system-of-systems (SoS).

The challenge

Quarry sites require a series of repetitive and sometimes dangerous tasks. A system of cooperating construction machines offers an opportunity to take humans out of harm’s way whilst providing a cost-effective and more environmentally friendly process for construction. But current safety standards and processes focus on single machines. In the case of cooperating autonomous machines, critical situations may be missed if safety assurance is applied for single machines, ignoring different application scenarios.

The research

The SUCCESS demonstrator project explored numerous aspects in the safety assurance of cooperating SoS, with a special focus on the construction machinery domain.

Specifically, the project:

• highlighted the challenges in hazard analysis of SoS 
• demonstrated a digital twin-based approach for simulation and verification of safety requirements
• evidenced the potential of geofences for dynamic risk management
• outlined an end-end tool framework for safety analysis

The project used a quarry site at Volvo Construction Equipment - the “Electric Site” - to give a real environment in which to study the fleet of machines working together as a SoS.

The results

You can read about all of the results from the project in the full project report. The main results include:

• the team proposed new ideas for the hazard analysis of SoS, including a state-based extension to System Theoretic Process Analysis (STPA) and the team’s “SafeSoS” approach. SafeSoS offers a hierarchical process for specifying a SoS and using those specifications for performing a safety analysis.
• the Volvo training simulators at Mälardalen University were adapted and extended to form a digital twin of the Electric Site. This enabled the team to verify the safety requirements of their collaborating autonomous systems and to implement and check multiple safety-critical scenarios.
• the project team has shown the potential of different types of geofences for dynamic risk management during the operational phase.
• an end-to-end tool framework was defined by the team that incorporates both process and product-based evidence, safety contracts, and dynamic assurance

Project partners

• Professor Sasikumar Punnekkat (principal investigator) Mälardalen University
• Professor Hans Hansson (co-investigator) Mälardalen University
• Stephan Baumgart (co-investigator) Volvo Construction Equipment
• Dr Henrik Thane (co-investigator) Safety Integrity AB