Putting safety on the map

Hey there, time traveller!
This article was published 08/04/2021 (1841 days ago), so information in it may no longer be current.

TWO University of Manitoba civil engineering professors will create maps of trucking routes on the Prairies, with the use of artificial intelligence, to make the network safer and more resilient.

Over five years, Jonathan Regehr and Babak Mehran will collect data on the number, type and weight of trucks that using the highways. Sensors will be embedded in roads in Manitoba, Saskatchewan and Alberta, along with parts of the northern territories.

They will also collect data on how traffic operates on the roads under different weather conditions. The scheme will be paid for using $1.6 million in research and development funding from the National Research Council of Canada.

Then, the researchers will work with Manitoba Infrastructure, the research council and private transportation software company International Road Dynamics to create models of the road network using artificial intelligence.

“The prairie road network is very large and sparse. We don’t have enough sensors to cover every element of it, so it’s a large network with missing data, for some lengths we have some data, for some we have nothing. One way to use (artificial intelligence) is to produce the information from existing data (through extrapolation),” Mehran said.

“So after identifying the most important roads and links in the Prairies, we form a model only based on those — AI will help us build an abstract network based on what we have.”

Researchers will be able to identify crucial links, and their weak points, in the logistical network. The data will reveal how different factors, human-made or natural, could create problems on the roads. The models will be updated when new data is added.

The data could be used to inform government decisions on infrastructure spending including upgrades to roads or bridges.

— Erik Pindera