Uncertainty Representation in a SOTIF-Related Use Case with Dempster-Shafer Theory for LiDAR Sensor-Based Object Detection

TL;DR

This work addresses uncertainty in LiDAR-based object detection within a SOTIF-related use case for automated driving. It adopts Dempster-Shafer Theory (DST) to construct a Frame of Discernment $\Theta$ with four outcomes, derive conditional BPAs from interdependent uncertainty sources, and resolve conflicts using Yager's Rule, while computing Bel and Pl for each hypothesis. A variance-based sensitivity analysis (VBSA) identifies the most influential uncertainties—particularly rain intensity, surface type, and LiDAR sensor performance—and guides mitigation strategies such as adaptive detection and expanded extreme-weather training data. The study demonstrates a structured approach to uncertainty representation in ADS perception, provides a pathway for risk-informed mitigation, and discusses limitations like BPA subjectivity and the static nature of the analysis, pointing to future work in dynamic, temporally-aware uncertainty propagation and integration with temporal models.

Abstract

Uncertainty in LiDAR sensor-based object detection arises from environmental variability and sensor performance limitations. Representing these uncertainties is essential for ensuring the Safety of the Intended Functionality (SOTIF), which focuses on preventing hazards in automated driving scenarios. This paper presents a systematic approach to identifying, classifying, and representing uncertainties in LiDAR-based object detection within a SOTIF-related scenario. Dempster-Shafer Theory (DST) is employed to construct a Frame of Discernment (FoD) to represent detection outcomes. Conditional Basic Probability Assignments (BPAs) are applied based on dependencies among identified uncertainty sources. Yager's Rule of Combination is used to resolve conflicting evidence from multiple sources, providing a structured framework to evaluate uncertainties' effects on detection accuracy. The study applies variance-based sensitivity analysis (VBSA) to quantify and prioritize uncertainties, detailing their specific impact on detection performance.

Figures (7)

• Figure 1: Workflow for representing uncertainty in a SOTIF-related Use Case
• Figure 2: SOTIF-related Use Case: Ego-Vehicle interacting with a cyclist and oncoming truck in adverse weather.
• Figure 3: DAG representing dependencies among uncertainty sources in a SOTIF-related Use Case for LiDAR-based object detection
• Figure 4: Workflow for applying DST in a SOTIF-related Use Case
• Figure 5: Conditional Probability distributions for detection outcomes across various sources of uncertainty