On-route incident prediction
Domain: Transport & Logistics
There are several aspects which could provoke a traffic incident, being the weather conditions one of them. This challenge aims to produce a prediction model which states when and where a traffic incident is going to happen, using weather data, past incidents data and social media data (for predicting traffic flows produced by mass events).
- Dataset 1: Incident dataset (RACC or mCLOUD)
- Dataset 2: Weather re-analysis and forecast data (UBIMET)
- Dataset 3: Events gathered from social networks or local event agendas as the ones from Open Data Euskadi or Bilbao Open Data.
- Dataset 4: Climate indicators for car traffic accidents in NRW, Germany (UBIMET)
The expected outcome of this challenge is to predict traffic incidents in order to allow road-assistance companies like RACC to offer better service.