Agung Nugroho Pramudhita, Meyti Eka Apriyani, Irsyad Arif Mashudi, Odhitya Desta Triswidrananta
Traffic accidents in Indonesia resulted in 28,131 fatalities from 139,258 cases in 2022 alone, highlighting a critical need for improved safety measures. This research aims to develop an advanced technology-based tool to enhance traffic accident detection and response. The proposed system utilizes a Raspberry Pi in conjunction with various sensors, including GPS, gyroscope, and accelerometer, to accurately detect accidents. Upon detection, the system promptly sends detailed accident information via a Telegram bot, ensuring timely communication with emergency responders. To further enhance the detection accuracy, the K-Nearest Neighbor (KNN) algorithm is implemented. The algorithm demonstrated significant improvements in accuracy, achieving 93% accuracy with K=3 and 90% with K=5, compared to 80% accuracy without the use of KNN. These results indicate that the incorporation of KNN substantially enhances the system's reliability in identifying accidents. Field trials underscored the system's effectiveness, as it successfully transmitted accident information through the Telegram bot without delay. Additionally, the data collected by the system is displayed on an internet-accessible website dashboard, providing a comprehensive overview of the accident's details. This dashboard allows for real-time monitoring and analysis, facilitating quicker and more informed decision-making by authorities. Overall, this development addresses the existing gap in accident detection technology, aiming to reduce the number of fatalities and improve response times in traffic accident scenarios. By leveraging the power of IoT and machine learning, this research contributes to the advancement of safer and smarter transportation systems in Indonesia. © 2025 Author(s).
State Polytechnic of Malang, East Java, Malang, Indonesia
Research at a Glance
Register to unlock
Topics & SDG Alignment
Register to unlock
Collaboration
Register to unlock
Author Profile (Selected)
Register to unlock
References Overview
Register to unlock
Journal & Source
Register to unlock
Metadata & Integrity
Register to unlock