Time loss is less than safety gain, according to University of British Columbia researchers

A new algorithm from researchers at the University of British Columbia shows the safest route to a destination. Integrated into a navigation device, the software guides drivers past accident-prone areas. Research leader Tarek Sayed and his graduate student Tarek Ghoul designed the new approach using real-time accident risk data. The algorithm can be integrated into navigation apps such as Google Maps, the researchers say.

Drones for analysis

The USB researchers used ten drones that hovered over downtown Athens for several days, recording the positions of vehicles, their speeds and acceleration values. This information has subsequently been used to identify near-misses between vehicles and then predict the risk of accidents in real time.

“This study is the first to use real-time accident risk data to provide navigation instructions and prescribe the safest route through a city. The algorithm can adjust course in real time and suggest detours to avoid dangerous locations. This will help increase road safety for all users. For example, companies will be able to navigate their fleet efficiently, prioritize safety and reduce the risk of accidents,” Sayed said. The prerequisite, he said, is that drones record traffic events around the clock.

Road a little longer

It turns out that the fastest routes are rarely the safest. For example, the team analyzed a small section of Athens’ road network and found that only 23 percent of the fastest routes were also determined to be the safest routes. “The safest route tended to be 22 percent safer than the fastest route, while the fastest route was only 11 percent faster than the safest route. This suggests that significant safety gains are achieved on the safest routes with only a small increase in travel time,” Ghoul says.