Back in 2019 and in cooperation with a leading Danish Insurance company, Cosesy installed more than 200 IoT and AI based early warning water monitoring systems in private homes.
Tons of data recorded over more than 24 months of operation have been analyzed and show these great results:
From year 1 to Year 2:
- Average homes have reduced the number of large water spills and damaged water pipes by 50% and
- Average homes have reduced the number of water leaks related to dripping toilets, showers and taps by 25%
As a general observation from the two years of operations it appears that:
- During summer, homeowners seem to be more relaxed as it takes longer to take actions on warnings. People may forget to turn off water when watering the plants, the poolmay accidentally flow over, or you are may be away from home when an accident occur.
- When rain and cold kicks in, people are more responsive to nudging messages from Cosesy’s leak detector and fix the problem before it develops into a large and costly damage.
Developed and released in 2019 the leak sensor was a "baby-AI" two years ago, and initially a bit too eager nudging our users. Over time, the sensor and our AI has learned much more about the individual homes and the general challenge of leak detection. Developing and optimizing algorithms and trained our pattern recognition on huge amount of data combined with feedback from users has grown our AI to a competent grown up.
Next up, we are also able to warn on risk of legionella and risks of frozen pipes.