Home/ ECE/ Explainable AI (XAI) for Automotive Radar
ECE · Seminar 10 · Trustworthy perception for safety systems

Explainable AI (XAI) for Automotive Radar

Explainable AI makes radar-based perception in vehicles interpretable, so engineers and regulators can understand, validate and certify the decisions of safety-critical detectors.

XAIautomotive radarADASinterpretabilitysafety

Modern ADAS and autonomous vehicles fuse cameras, LiDAR and radar — radar being uniquely robust in rain, fog and darkness, and able to measure velocity directly via Doppler. Deep networks now classify radar returns, but they are black boxes. For a safety-critical system that must be certified (ISO 26262 / SOTIF), engineers need to know why a detection was made — that is the role of Explainable AI (XAI).

Working principle

Radar produces range–Doppler–angle data cubes that a CNN turns into object detections. XAI methods add an interpretation layer: saliency / Grad-CAM highlights which regions of the range–Doppler map drove the decision; SHAP attributes the output to input features with game-theoretic consistency; surrogate models approximate the network locally. The explanation lets developers spot spurious cues (e.g. clutter) and build the safety argument.

Radar data cube1CNN detector2Detections3XAI (Grad- CAM/SHAP)4Human-readable reason5Adding an explanation layer to a radar perception model
Figure 1. The XAI stage exposes which range–Doppler features the network relied on, supporting validation, debugging and certification.
Table 1. Common XAI techniques for radar perception
MethodTypeWhat it shows
Grad-CAMSaliency mapInfluential input regions
SHAPAttributionPer-feature contribution
LIMELocal surrogateLocally faithful explanation
AttentionBuilt-inWhere the model focuses
CautionExplanations must be faithful to the model, not just plausible to humans. A convincing-but-wrong explanation is dangerous in a safety case — evaluating explanation fidelity is itself a research problem.

Applications

  • Validating ADAS object detection and false-alarm analysis
  • Safety certification evidence (ISO 26262 / SOTIF)
  • Debugging sensor-fusion and corner-case behaviour

References & further reading

  1. Selvaraju et al., “Grad-CAM: Visual Explanations from Deep Networks,” ICCV 2017.
  2. Lundberg & Lee, “A Unified Approach to Interpreting Model Predictions (SHAP),” NeurIPS 2017.
  3. ISO 26262 / ISO 21448 (SOTIF) functional-safety standards.