In each room, we install two devices: one is a sensor, and the other is a computer that can understand what the sensor sees. The video from the sensor is then converted into small stick figures and text, which become available to you on mobile and computer/tablet.
A small example of how the sensor understands the world. In the video, you can see a resident lying in a bed and two staff members helping them. The squares are other things that the sensor can recognize, such as beds or chairs.
All images from the sensor are deleted as soon as the computer has viewed and converted them into stick figures (usually after 15 seconds). This way, we do not need to store any images for the system to function.
There is only one case where we keep images for longer than 15 seconds: if we have received explicit consent to improve the system. Our technology only gets better at understanding its surroundings if we provide it with real video examples. This is why we often seek consent during the start-up phase in a new ward. This allows us to train the sensor to better understand the new rooms and areas where it is installed.