A Framework for Pothole Detection via the AI-Blockchain Integration


Mubarak A. S., Ameen Z. S., Tonga P., Altrjman C., Al-Turjman F.

2nd International Conference on Forthcoming Networks and Sustainability in the IoT Era (FoNeS-IoT), ELECTR NETWORK, 8 - 09 January 2022, vol.129, pp.398-406 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume: 129
  • Doi Number: 10.1007/978-3-030-99616-1_53
  • Country: ELECTR NETWORK
  • Page Numbers: pp.398-406
  • Keywords: Artificial intelligence, Blockchain, Computer vision, EfficientDet, ASPHALT PAVEMENT
  • Istanbul University Affiliated: No

Abstract

Potholes are a major cause of road accidents. In smart cities, developing a system that will help in detecting the potholes in real-time while warning the motorist before reaching that location is very important. Lightweight EfficientDet family models were proposed in the literature to detect potholes in real-time. The work is divided into two parts, first a mobile application was developed using these trained models and it can efficiently detect the potholes. Also, in the second part, a blockchain was proposed to keep a record of the location of the pothole in real-time so that it can be updated on the network. The blockchain serves as a security mechanism so that the potholes coordinates can be altered and become up to date.