From research to reusable engineering patterns

Human-in-the-Loop Safety for ADAS: Blending Kansei, HSI, and AI to Reduce False-Positive Braking

Author: Nuri Başer Üstün, CSEP — DNB Engineering

The study argues that false-positive braking is not solely a perception accuracy problem. Driver intent, trust, workload and human-machine interaction must be considered together.

Proposed approach

The “Nuisance Guard” functional pattern adds threat consensus, driver intent arbitration and explainability hooks to the safety-critical decision path, supervised by deterministic envelopes and runtime monitors.

Transfer potential

The approach can also support human-supervised defense autonomy, remote operation and decision-support systems.