Sensor Degradation Detection Algorithm for Automated Driving Systems

Darab, Jonathan M; Witcher, Christina J · 2023 · ROSA P / Safety through Disruption (Safe-D) University Transportation Center (UTC)

archive: archived pipeline: cataloged verified

Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)

Summary

This research addresses the critical safety challenge of sensor degradation in Automated Driving Systems (ADS), where environmental factors like weather or cyberattacks can corrupt sensor data, leading to hazardous vehicle behaviors such as off-road excursions or sudden stops. The study aimed to develop and evaluate a sensor degradation detection algorithm capable of identifying misinformation in LiDAR, radar, and GPS signals. The project was a collaborative effort involving the Virginia Tech Transportation Institute (VTTI), Old Dominion University (ODU), and the Global Center for Automotive Performance Simulation (GCAPS). The methodology relied on a virtual simulation framework built using MATLAB and IPG CarMaker. Researchers selected 100 crash and near-crash events from the SHRP 2 Naturalistic Driving Database to establish baseline performance metrics, expanding these into 1,000 simulated events. GCAPS developed phenomenological sensor models for LiDAR and radar based on empirical data collected from the Virginia Smart Roads, while ODU developed GPS models and the detection algorithm. The detection system utilized the DeepPOSE framework, which employs a combination of convolutional neural networks and sequence-to-sequence models to estimate vehicle position from inertial measurement unit (IMU) data and detect GPS spoofing. For LiDAR and radar, the team converted sensor signals into time-frequency representations (scalograms) and used a GoogLeNet backbone to classify baseline noise versus degradation noise. The results indicated that the developed detection algorithm achieved 70% accuracy in identifying degraded sensor states. The study successfully demonstrated the feasibility of using simulation and deep learning to detect sensor misinformation, including effects from rain intensity and GPS random variation. However, the authors concluded that the current accuracy level is insufficient for immediate implementation in vehicle systems. The research highlights that additional training methods and algorithmic adjustments are necessary to meet the reliability standards required for operational ADS. The work establishes a foundational process for collecting sensor data, creating statistical sensor models, and utilizing simulation for algorithm development, contributing to the broader field of ADS safety and cybersecurity.

Key finding

The developed sensor degradation detection algorithm achieved 70% accuracy in identifying degraded sensor performance within a simulated environment.

Methodology

simulator

Sample size: 100

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• telematics crash prediction

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Methodological Resource: validation psychometrics