Allocate ambulances at different city locations in the most efficient way possible. Historical information of accidents combined with weather conditions shows that certain locations of the city are more likely to observe car accidents.


The team built a classification model to predict the probability of a car accident occurring at the zip code level. The trained model considers historical car accidents and weather conditions to predict accidents probability. The probabilities for each zip code are used to feed a decision optimization model that outputs the optimal ambulance locations. s demographics like country and industry among others.


  • Government

Tools Used

  • Data Science Experience (DSX)

Data Science Techniques

  • Binary Classification, Decision Optimization


The combination of using machine learning to predict accidents and decision optimization to locate ambulances given the accidents probabilities ensure an optimal way to allocate ambulances. As a result, ambulances can potentially get to the accident locations faster than before.

Need help with a similar case?

Contact us to find out how we can help!