본문 바로가기
  • Home

Real-Time AI Inference Method for Verification of Vehicle Sudden Unintended Acceleration (SUA)

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2025, 30(11), pp.71~78
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science
  • Received : October 10, 2025
  • Accepted : October 29, 2025
  • Published : November 28, 2025

HeeSeok Choi 1

1주식회사 엘가나

Accredited

ABSTRACT

In recent years, the number of sudden unintended acceleration (SUA) incidents has been steadily increasing, and research and discussions are actively taking place in many countries, including Korea. Moreover, as drivers involved in accidents caused by pedal misapplication often claim SUA by default, there is a growing need for solutions that can clearly distinguish between technical failure and pedal misapplication. To address this need, this paper presents a foundational study on a cloud-based AI solution for the verification of SUA incidents from the consumer perspective. Specifically, the proposed solution measures the displacement of the brake and accelerator pedals at the exact moment of SUA. In addition, a pedal camera is installed to monitor the driver’s pedal operation, while real-time AI inference results are recorded and stored. As a result, the findings of this study are expected to contribute to the development of more reliable SUA verification systems. Furthermore, the proposed approach can serve as a fundamental dataset for improving driver pedal image data processing in future research.

Citation status

* References for papers published after 2024 are currently being built.