EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Smart Accident Analysis: An Automated System for Determining Car Accident Fault Responsibility

Huda Al‐Muzaini, Sarab Almuqati, Shaykhah Aldawsary, Mashael Zarea, Nada Alsultan

IET Intelligent Transport Systems · 2026-01 · 20 (1)

DOI: 10.1049/itr2.70287 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IET Intelligent Transport Systems, online erschienen am 21.08.2026. Verfasst von Huda Al‐Muzaini und 4 weiteren Autor:innen. Die Arbeit nennt 25 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Car accidents are a frequent occurrence in modern urban environments, often followed by disputes regarding fault attribution. Traditional methods of determining responsibility rely heavily on human judgment, which can be time‐consuming, inconsistent and subject to bias. This paper introduces an AI‐powered system designed to analyse car accident descriptions and determine fault responsibility automatically. Leveraging advanced natural language processing techniques and deep learning models, the system processes unstructured textual reports submitted by both parties involved in an accident. It extracts key incident details and delivers fault percentage assessments based on contextual understanding. The study formulates fault estimation as a direct prediction of bilateral fault‐percentage splits under a constrained responsibility framework (0/25/50/75/100). A Saudi‐aligned labelling protocol combining expert input with controlled GPT‐assisted annotation is introduced to ensure consistency with local traffic‐assessment practices. Furthermore, a systematic comparison of three modelling strategies: flat, dual‐model and hierarchical across multiple deep‐learning backbones is conducted to identify the optimal configuration. Among the models evaluated, RoBERTa demonstrated the highest performance, achieving an overall F 1 score (the harmonic mean of precision and recall) of 94%, significantly outperforming baseline approaches. This solution not only enhances the efficiency and fairness of accident investigations but also represents a step forward in the application of AI in legal and insurance‐related domains.

Zitieren

APA

Al‐Muzaini, H., Almuqati, S., Aldawsary, S., Zarea, M., & Alsultan, N. (2026). Smart Accident Analysis: An Automated System for Determining Car Accident Fault Responsibility. IET Intelligent Transport Systems, 20(1). https://doi.org/10.1049/itr2.70287

Vancouver

Al‐Muzaini H, Almuqati S, Aldawsary S, Zarea M, Alsultan N. Smart Accident Analysis: An Automated System for Determining Car Accident Fault Responsibility. IET Intelligent Transport Systems. 2026;20(1). doi:10.1049/itr2.70287

BibTeX
@article{AlMuzaini2026,
  author = {Al‐Muzaini, Huda and Almuqati, Sarab and Aldawsary, Shaykhah and Zarea, Mashael and Alsultan, Nada},
  title = {Smart Accident Analysis: An Automated System for Determining Car Accident Fault Responsibility},
  journal = {IET Intelligent Transport Systems},
  year = {2026},
  volume = {20},
  number = {1},
  doi = {10.1049/itr2.70287}
}
Kostenlos starten →