Problem: Lack of knowledge of the conditions of the sewer system leads to premature replacement of pipes costing the Danish society approx. 173 million DKK per year and 10.7 billion per year for the OECD countries.
Solution: A robot that uses computer vision and AI to automatically inspect the sewer system and report to a dynamic asset management system.
Duration: 2018 – 2022
Funding: Innovation Fund Denmark
Budget: 23.8 mio. DDK
The project content:
The goal of ASIR is to develop a working prototype of an automatic sewer
inspection system as illustrated below. The goal will be reached through three objectives:
1) To develop an autonomous robot to allow much more frequent, efficient and safe
inspection of the otherwise difficult-to-access areas inside sewage pipelines. A
small and lightweight robot platform that can carry a sensor suite, lighting, battery
and onboard processing capability. The robot will include an algorithm that can
navigate and control the robot platform in pipes with a diameter down to 20cm.
2) To develop an AI vision solution for automatic diagnostics and annotation of the
condition of a pipe system based on video from the robot. An algorithm capable of
detecting and classifying the abnormalities in a pipe is expected to perform better
than the traditional manual inspection.
3) To develop an asset management system with Big Data to use the frequent
inspection data with a time dimensional analysis and with GIS and other data to
understand and minimize asset repair and maintenance costs. ASIR will result in a
paradigm shift within the field of pipe inspection. This, in turn, opens up new
processes for forecasting maintenance of pipelines.