In Norway, 26% of fatal accidents in 2017 were attributed to vehicle technical issues, with poor tire conditions frequently cited as a contributing factor . This causes avoidable fatalities year after year, highlighting the potential for road authorities to identify problematic trucks proactively. Norwegian company Norsk Datateknikk is now aiding the government in increasing road safety in their latest project. They are what one would call „real full stack developers“: they code, but also engineer and design the electronics and mechanical solutions. Now they are testing to install camera sensors on Norwegian roads to check truck tires with the help of AI.
For the project, eight cameras are built into the concrete structure of the road at a control station south of Trondheim. They operate at 90 fps to catch every passing vehicle while EdgeKit is waiting on the side of the road to handle incoming images in real-time. EdgeKit specializes in Edge Computing, which ensures data security as well as very quick results without latency issues due to shorter data paths. Pictures that don’t show tires of trucks are thrown out immediately, the rest is categorized by a neural network to identify things like summer and winter tires, spikes or no spikes, way too worn tires, or one-sidedly worn tires which indicate a disproportion in air pressure. After the categorization and therefore identification of trucks with a high-risk factor, drivers in question are asked to pull over for a further check-up. This way not all vehicles have to be checked all the time to ensure safety.
EdgeKit is for some months now sitting in a heated locker, safe from snow, ice, and thunderstorm, at the side of the road, working through development and testing. Norsk Datateknikk reported having searched for industrially usable microcomputers for their case but found it challenging to easily compare pricing as no other company was able to provide final pricing immediately.
EdgeKit on the contrary offers an online live pricing view for all possible configurations as well as continuous support before choosing an individual solution and during and after deployment.
Do you have any questions regarding this project or our technology? Send us an e-mail! We’ll reply asap. firstname.lastname@example.org