Installing sensors around the house which notify homeowners of water leaks or fire is no longer enough. When machines have the ability to learn, they may even predict disasters before the unfortunate actually happens.
Installing sensors around the house which notify homeowners of water leaks or fire is no longer enough. When machines have the ability to learn, they may even predict disasters before the unfortunate actually happens.
A startup has announced a home protection system called OnePrevent claimed to have the ability to detect household disasters such as fires and floods 20 minutes before their occurrence. This will give users extra time to act in such situations.
The OnePrevent system includes smoke and carbon monoxide detectors, heat and humidity sensors, motion sensors and open/closed detectors attached to door or windows.
The core system is powered by a custom-built AI algorithm called OneEvent. According to the company, the cloud-based analytic engine needs a few weeks to learn what normal conditions are like around the home, before it is able to recognize abnormal situations like fires and break-in.
The algorithm aggregates data from all users to learn from it. The idea is that after more and more people install OnePrevent sensors, the system will get smarter and smarter. For example, it will be able to tell the difference between a burned piece of toast and a building on fire.
The startup has partnered with local fire officials in Wisconsin, United States. By installing OnePrevent sensors in homes, the company trains its algorithm to recognize when a fire may be about to break out.
However, compared to usual fire or flood alarm systems, One Prevent carries a higher price tag. The whole kit costs US$299 plus a $25 monthly service subscription fee.
It’s not the cheapest option for homeowners, but it’s interesting to see how companies are using machine learning and artificial intelligence technologies to make homes become aware of living environment and make predictions about them.