Safety analysis and verification/validation of MachIne LEarning based systems

Purpose and goal:

Develop enabling technologies and safety assurance framework for transport services enabled by small autonomous vehicles ("AMR"): transport robots and delivery robots.

The main objective of SMILE IV is to transfer vehicular safety practices from SMILEI/II/III into the innovative transportation services facilitated by AMRs, with a special focus on two typical use cases: autonomous transport robots in industrial environments and delivery robots in public spaces. Expected results and effects:

Facilitate the adoption of near future transport and mobility services enabled by connected small autonomous vehicles by:

developing safety strategy&architecture for connected AMR services; applying safety assurance process and safety supervisor architecture for timeseries data; developing anomaly detection algorithms for timeseries data; and identify approaches can be taken to create a data management strategy that seamlessly integrates synthetic data generated by simulators with real-world captured data, while ensuring safety compliance throughout the AI lifecycle.

Approach and implementation:

The project is planned for 39 months (Q3/2023-Q4/2026) with seven inter-correlated work packages and two use cases: Transport robot in factory settings & Delivery robots in public spaces. The project consortium consists of: RISE ( coordinator), Volvo Group, Dynorobotics, Infotiv, and Halmstad University.