CITY OF MEMPHIS
In an emergency : Memphis Fire Department and ( left )
Memphis Paramedics in action potholes . Previously , the city would rely on tips from members of the public , responding by sending public employees to resurface problem roads . As with many use-cases , AI is allowing this manual process to be completely overhauled .
Instead , Memphis has been tasking its public works and information technology departments with finding a technological solution to potholes . They joined forces with Google and SpringML , who suggested mounting panoramic cameras on the front of city buses . The cameras create a 360 ° image of their surroundings along bus routes – which , after all , tend to follow the city ’ s most heavily used arterial routes – and an AI algorithm helps to detect what is a pothole and what is just a change in the appearance of the road surface .
Initially , 30 days of raw footage was captured to help train the algorithm ; and now the system automatically creates and assigns repair tickets to available crew members , who are then dispatched to refill the road . Not only did this make detection quicker and more effective , it has cut down on the number of man hours involved – partly because crews are no longer required to go out and verify repair requests from members of the public .
“ Really , Memphis was one of the first cities in the US to do this as a proof of concept ,” Boateng explains . “ It worked out so well that we are now going to expand it .” This means evaluating more criteria – for example , could the system be used to detect abandoned buildings and blight – as well as increasing the amount of video capture that takes place . As well as public buses , the city has access to CCTV cameras and even dash cameras operated by the police department ; these already have AI capabilities built into them , as they ’ re used to automatically detect fugitive number plates . Could this
cybermagazine . com 123