“A fully autonomous pothole detector that updates a visual backend in real-time, and also provides detailed data stored on an SD card.
Each year, Renault hosts a contest of hacking its Twizy platform (a design based on an electric urban vehicle). We participated in that contest (via our university, Polytech Sorbonne) and came up with the idea of hacking the platform to make a pothole detection vehicle, that can update a map with the positions of the potholes in real time. Here’s the promotional video we made about it (there are subtitles in English):
So how does it work?
Its array of 6 infrared sensors mounted under the vehicle pointing downwards detect changes in distance between the vehicle and the road, thus detecting potholes. The system’s accelerometer can also detect the potholes and abrupt shocks the vehicle experiences due to abnormal vertical variations.
The system then sends that information (via the Sigfox module) to a backend that we made using Microsoft Azure, and then the information is displayed on a map interface we made using Power BI. The system also stores detailed data on an onboard micro-SD card in JSON format, so the user can recuperate that information in order to construct a clearer image of the severity of each pothole the system has detected.”