As the world becomes more urbanized, improving our cities becomes more important than ever. AI holds the key to making cities more efficient, safe and green. But to achieve these goals at city scale and country scale, AI must be deployed in a distributed manner. NovuMind’s technology enables AI to be embedded throughout the fabric of the modern city, in buildings, in transportation links, and other infrastructure.
NovuMind has worked with Norwegian Public Roads Administration (NPRA) to enable AI-powered bicycle counting in the area of Prinsens Gate in Trondheim, the third biggest city in Norway. NPRA was looking for a way to measure the success of bicycle transportation programs. They wanted a system that could count bicyclists on the road efficiently, accurately, and without violating citizens’ privacy. The solution was a computer vision system using NovuTensor technology embedded in roadside cameras. The system identifies and counts bicycles with an accuracy rate of 96.4%, which is better than human observers could achieve. To preserve riders’ privacy, no images are stored or transmitted by the cameras. Only anonymous statistics are sent to the cloud. This solution increases scalability and preserves citizens’ privacy.