Detektering av igentäppta dagvattenbrunnar i smart cities

Sammanfattning: In the urban society there are more and more hard surfaces that prevent the natural drainage of rain water. To lead the runoff of rain water and to avoid flooding it is crucial that the systems for drainage work properly. In the event of a storm drain being blocked, flooding can occur due to the rain water not being able to flow in to the storm drain. A system that can send an alarm when a storm drain is blocked before flooding has occurred is a step towards so-called smart cities. To prevent flooding, this bachelor thesis has explored how such a system can be designed to detect and send an alarm when a storm drain is blocked, e.g. by leaves and debris. The proposed system measures the level and the expansion of a body of water together with the intensity of rain. The measurements are then used in a mathematical model which will detect a fault, that is, a storm drain not being able to let water through when it is supposed to. The measurement of the level of rain water is carried out by a radar, and the expansion is measured with a camera which utilizes image segmentation. The intensity of rain is monitored by the company Waterview’s solution that uses a camera as a rain sensor. The bachelor thesis is written at Axis Communications AB in Lund.