Dynamic Risk Assessment Methodology with an LDM-based System for Parking Scenarios

TL;DR

The paper addresses risk-aware ADAS in parking by integrating interior driver state with exterior scene data using a Local Dynamic Map. It defines a dynamic risk methodology with a TTC/awareness-based risk scale and risk zones, and creates a multi-sensor dataset for benchmarking. It introduces a Local Dynamic Map–based Dynamic Risk Assessment System (DRAS) with interchangeable detectors (LiDAR-based pedestrian detection and gaze detection) and demonstrates feasibility on the dataset. The study achieves an overall risk-assessment accuracy of about 83% and discusses extensions to larger vehicles, full-surround risk estimation, and enhanced detection algorithms.

Abstract

This paper describes the methodology for building a dynamic risk assessment for ADAS (Advanced Driving Assistance Systems) algorithms in parking scenarios, fusing exterior and interior perception for a better understanding of the scene and a more comprehensive risk estimation. This includes the definition of a dynamic risk methodology that depends on the situation from inside and outside the vehicle, the creation of a multi-sensor dataset of risk assessment for ADAS benchmarking purposes, and a Local Dynamic Map (LDM) that fuses data from the exterior and interior of the car to build an LDM-based Dynamic Risk Assessment System (DRAS).

Figures (7)

• Figure 1: Diagram of different scenarios in a parking maneuver
• Figure 2: Diagram of risk zones
• Figure 3: Risk levels scale
• Figure 4: Risk assessment dependant on distance and driver awareness
• Figure 5: Images of the recording process of the dataset (indoor and outdoor scenarios)