Collision (or crash) Severity Inference

Geotab is a global leader in IoT and connected transportation, providing valuable insights to over 80,000 customers worldwide by collecting more than 4 billion data points daily. The Safety & Video Analytics team at Geotab currently detects vehicle collisions using in-house machine learning models on telematics data. However, the severity of crashes is currently assessed solely through g-force measurements, which is not always reliable.
This project aims to develop a more sophisticated method to infer crash severity using additional data sources, such as kinematic data, contextual data lakes, and external crash databases (e.g., Federal Motor Carrier Safety Administration – FMCSA). The new methodology will help in collision reconstruction, insurance claim processing, and overall vehicle safety analysis. By providing a more precise and data-driven severity assessment, Geotab can enhance its services for customers and insurance partners, leading to better risk evaluation, pricing, and claims management. The success of this project would help Geotab provide more contextual information about a crash to the customers that can be used for planning and for litigation purposes. This will also help us build more strategic partnerships by providing more value to our insurance partners

Faculty Supervisor:

Andrei Badescu

Student:

Partner:

Geotab Inc

Discipline:

Computer science

Sector:

Information and cultural industries; Professional, scientific and technical services; Transportation and warehousing

University:

University of Toronto

Program: