Cause-and-Effect Analysis of ADAS: A Comparison Study between Literature Review and Complaint Data

Ayoub, Jackie; Wang, Zifei; Li, Meitang; Guo, Huizhong; Sherony, Rini; Bao, Shan; Zhou, Feng · 2022 · Unknown

DOI: 10.1145/3543174.3547117

archive: archived pipeline: cataloged verified

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

Summary

This study investigates the causes and effects of Advanced Driver Assistance Systems (ADAS) failures by comparing findings from academic literature with real-world consumer complaint data. The research is motivated by the need to understand ADAS limitations to improve vehicle safety and customer experience, noting that improper use or distrust can negate safety benefits. The authors aimed to identify discrepancies between academic focus areas and actual user experiences to provide targeted recommendations for ADAS improvement. The methodology involved two parallel analyses. First, a literature review was conducted on 67 relevant papers published between 2016 and 2021, categorized by limitations and proposed solutions. Second, the authors analyzed 440,999 consumer complaints from the National Highway Traffic Safety Administration (NHTSA) database. Using a Bidirectional Encoder Representations from Transformers (BERT) classifier, they identified 1,141 ADAS-related complaints. To extract cause-and-effect relationships from these texts, the team developed a natural language processing model using MPNet, fine-tuned on a manually annotated dataset of 542 sentences combined with two public datasets. The causal extraction model achieved an F1-score of 0.842, outperforming BERT and ELECTRA models. The results revealed significant divergences between academic research and consumer complaints. While both sources identified human, environmental, and vehicle factors as primary causes, academic literature focused predominantly on human factors (40.7% of limitation studies), proposing algorithmic solutions for driver monitoring and behavior prediction. In contrast, consumer complaints were dominated by vehicle factors (71.2%), with the top specific causes being false alarms (17.3%), failure to respond (13.1%), and sensor issues (5.5%). Environmental factors accounted for only 5.0% of complaints, and human factors for just 1.2%. The most frequent effects reported by consumers were ADAS failure (18.8%), hard braking (13.0%), and warning alerts (10.5%). The study concludes that academic research and consumer data offer complementary insights. Academic work heavily addresses human-centric issues and algorithmic improvements, whereas consumers primarily report hardware and software malfunctions, such as false alarms and sensor errors. This gap suggests that current research may overlook the technical reliability issues that most impact user trust and satisfaction. The findings imply that future ADAS development should integrate algorithmic advancements with robust hardware and software reliability to address the specific failures driving consumer dissatisfaction.

Key finding

Academic research focused more on human factors and algorithmic solutions for ADAS issues, while consumer complaints were dominated by vehicle-related failures such as false alarms and sensor problems.

Methodology

mixed_methods

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 author_sweep_intake on 2026-05-28.

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.

• adas effectiveness