Automatic identification of defects for sewage pipes

Together with the Swedish Tekniska verken, we developed a computer vision-based solution that learns from each defect, and continuously improves the level of automation.

Tekniska verken is a municipal company that works with 260,000 private and business customers by offering a wide range of products. They include electricity, lighting, water, district heating, district cooling, waste management, broadband, biogas, efficient energy solutions as well as electricity trading.

We built a robotics and computer vision-based solution that helps Tekniska verken in their labor-intensive task to monitor blockages, leaks and foreign objects within sewage pipes monitored via video.

Intelligent awareness of foreign objects

The computer vision-powered solution analyzes live video and flags the defects to a human to be verified – empowering Tekniska verken’s employees with intelligent awareness of foreign objects and potential blockages in sewage pipes. The solutions enables faster, more efficient and reliable identification of defects.

The AI solution, currently in use at the Swedish municipality of Linköping, leverages various computer vision models that aim at identifying objects in the sewage pipes. The collaboration between the two companies enables more efficient and reliable sewage pipe management, thus contributing to Tekniska verken’s goal of creating a more sustainable society.


  • Increased safety and reliability in the region
  • Modern AI-powered system that increases automation, thus permitting humans to focus on higher-value creating tasks
  • Enables faster response and reacting on time to any anomalies in the pipe network

Read the press release for more information.


Silo Operating Software used in this project

Annotation platform
Development platform
Operations platform
The AI solution increases safety and reliability of the pipe network in the entire region of Linköping. Computer vision technologies are particularly apt for improving monitoring and situational awareness as they can seamlessly support our employees in their laborious monitoring and maintenance work. Thanks to Silo AI’s expertise, we’ve been able to build a modern AI-powered system that meets today’s requirements.
Fredrik Danell
Business Developer at Tekniska verken

Interested in becoming
more AI-driven?

Image of Peter Sarlin

Peter Sarlin

CEO, Silo AI
+358 41 507 6066

You might be interested in

Climate change data extractor

Canadian scenario generation company Riskthinking.AI and Silo AI co-created an AI solution to automatically extract climate risk data. We built a natural language processing-powered solution, that brings out insights related to one consequence of climate change: the rise of sea level.

Read more

Read More
Contact Silo AI
Considering elevating your business with AI? Look no further. We recommend you to get in touch early on to get the best value of our expertise and collaboration.