Solutions for presence detection in an intercom door station

In this work, different solutions for human presence detection in a door intercom device is examined. The study was conducted at a company in Lund, Sweden. The aspiration was to have a reliable presence detection to be able to put the device in sleep mode when not used, and the current solution is not optimal because of limiting design requirements and false triggers. In the project, several different sensor technologies and sensors were examined and evaluated. After an initial selection, four different sensor technologies were studied and six different sensors using these technologies were investigated both theoretically and practically during tests. The sensor technologies investigated were passive IR, active IR, ultrasonic and radar. With the knowledge gained during the investigation and experiments, the different sensors suitability for the project were discussed and evaluated.

The technology most appropriate for this project was found to be radar, due to its ability to easily be installed and integrated inside the intercom since it can work through a lot of materials, including the hardened glass found on the device. While the first radar sensor that was tested demonstrated the potential of the technology, it was found to be a bit too simple. There are a lot of sophisticated radar sensors available but after much investigation the one ultimately chosen was the 60 GHz Acconeer A121 Pulsed Coherent Radar Sensor due to its compatibility with the specific use case. This sensor allows for a much smarter presence detection than the one currently used in the intercom. After researching and testing the sensor it was concluded that it could be integrated into the intercom’s PCB and function through the hardened glass panel on the front. This makes it much easier to implement than the currently used solution by having less limiting requirements on the mechanical design of the intercom. The sensor enables more efficient presence detection because of the ability to activate only when someone approaches, and filtering out unwanted triggers with the help of the sensor’s high accuracy.

Supervisors: Johan Björnstedt (IEA), Alexander Holmström (Axis)
Examiner: Fran Marquez (IEA)