Building the safety case for a robot that tests autonomous trucks
Client & context
AstaZero (RISE — Research Institutes of Sweden) operates the world's first full-scale independent test environment for automated transport systems. To validate SAE L3-L4 autonomous trucks, AstaZero developed a Test Robot that physically simulates the automated driving system on a closed proving ground.
The problem
Test equipment that drives a real truck must itself be provably safe — but ISO/TS 5083:2025 was brand new, the machinery aspects fell under ISO 13849, and the existing tooling could not produce the traceable requirements and timing evidence (FTTI) the safety argument demanded. Without a defensible safety case, validation campaigns for customer trucks could not proceed.
What I did
- Took ownership of all safety activities for the Test Robot: safety concepts, validation strategies, and the safety-case evidence, under ISO/TS 5083 and ISO 13849.
- Performed vehicle-dynamics and failure-mode analysis on the test track to verify safe behaviour of the validation equipment.
- Built (solo, AI-assisted) a full-stack Requirement Management web app — Fastify 5, PostgreSQL 16 with append-only history and immutability triggers, JWE session auth — making requirement traceability and safety-case evidence first-class citizens of the process instead of spreadsheet exports.
- Built a simulation tool computing Fault Tolerant Time Interval (FTTI) for vehicles on the proving ground, modelling vehicle dynamics and failure-response timing per ISO 26262 to underpin the safety concept.
The outcome
A managed, evidence-backed safety case for SAE L3-L4 test equipment; a requirement-traceability tool the safety process now runs on; and FTTI evidence derived from simulation rather than assumption.
Standards & tools
ISO/TS 5083:2025ISO 13849ISO 26262SAE L3-L4FastifyPostgreSQLVehicle dynamics simulation
Facing something similar?
This is the kind of work I do under AV & Autonomous Systems Safety.
Book an intro call