UBC Canada Researchers Introduce A New AI Algorithm That Maps Safest Routes For City Drivers

Navigation apps are applications that provide navigational directions in real-time. Most of the available navigation apps can show us the fastest possible traveling route to the designated location but these apps cannot tell us the safest route to our preferred destination. 

It is important to note that the quickest route does not always guarantee the highest level of safety. Given the paramount importance of safety, extensive research is underway worldwide to enhance navigation systems and enable the identification of the safest and most efficient routes.

The researchers at UBC have developed an algorithm that could navigate and suggest the safest possible route. The research team has developed a method that can identify the safest possible route in an urban network using real-time crash risk data. Further, this algorithm can also be incorporated into navigation apps such as Google Maps, making it accessible to everyone.

To facilitate this research, the team used 10 drones operating over the city of Athens, Greece, for multiple days and gathered data generated by these drones. The data they gathered include the vehicle position, speed, and acceleration. This information was very crucial to identify near-misses between vehicles, and then they predicted the risk of crashes between the vehicles in real-time. This research seeks to develop a real-time routing algorithm that considers the magnitude of crash risk along different points in a particular route and also the duration of exposure to those conditions. The safest routes are then compared to the fastest routes, and the trade-off between safety and mobility is examined. 

The results of this study are quite interesting. The study showed that the safest routes tended to be 22 percent safer than the fastest routes, but the fastest routes were just 11 percent faster than the safest routes. In many situations, the safest route algorithm tends to follow the same route as the fastest route, making a detour at a specific point to avoid what it identifies as a hazardous location. In fact, the safest route is the fastest route 54 percent of the time. The researchers said that road users should consider a mix of safety and efficiency when choosing directions. 

This experiment suggested that there is a trade-off between the fastest route and the safest route. Also, this study has paved the way for various other research areas of the domain. In the future, the safety of one route can also be quantified relative to the safety of other routes. 

The limitation of this model is that the data is being collected only for a city and only over a limited period of time and may not accurately describe the location of the traffic environment changes substantially. So, it would be beneficial to test this model on a much larger scale for a greater duration and more available routes to determine and potentially generalize the trade-offs between safety and mobility over long distances. Consequently, the researchers are currently extending the scope of their research to some other cities.


Check out the Paper and Blog. All Credit For This Research Goes To the Researchers on This Project. Also, don’t forget to join our 26k+ ML SubRedditDiscord Channel, and Email Newsletter, where we share the latest AI research news, cool AI projects, and more.

🐝 Join the Fastest Growing AI Research Newsletter Read by Researchers from Google + NVIDIA + Meta + Stanford + MIT + Microsoft and many others...