Search papers, labs, and topics across Lattice.
This paper introduces a Subjective Logic (SL)-based method for dynamically updating confidence in safety arguments using runtime Safety Performance Indicators (SPIs). The approach integrates design-time evidence with windowed runtime SPIs to quantify and propagate confidence across the development lifecycle. The method prioritizes responsiveness to safety violations by rapidly penalizing confidence when violations occur, demonstrating its effectiveness via a simulation of a construction zone assist function with an ML-based cone detection component.
Real-time safety assurance gets a boost: this method dynamically adjusts confidence in safety arguments based on runtime performance, reacting swiftly to violations.
We present a method for dynamic quantitative assurance that enhances static safety cases with continuous, runtime-driven confidence updates. The method quantifies and propagates confidence across the development lifecycle by integrating design-time evidence and windowed runtime Safety Performance Indicators (SPIs) within a single Subjective Logic (SL)-based assurance case. At runtime, SPI evidence is continuously evaluated, and targeted claims are updated using a rule that increases confidence in the absence of violations and imposes prompt penalties when violations occur. This design prioritizes safety-relevant responsiveness over exact classical Bayesian posterior updates. We demonstrate the method using a simulation-based construction zone assist function, focusing on an ML-based construction cone detection component, and show how confidence evolves as SPI evidence is observed in operation.
