CoronaSense, the automatic temperature sensor
An interview with Adrian Shajkofci, PhD student at EPFL
Portrait of Adrian Shajkofci
CoronaSense is an automatic temperature sensor developed by Adrian Shajkofci during containment. After obtaining a master’s degree in biotechnology at EPFL, he is back in Valais, and more precisely at the IDIAP Research Institute, where he is finalizing a PhD in electronics. Adrian likes to put his skills in computer science at the service of companies. We can say that Adrian is an inventor, since he designed an automatic temperature detection system during containment at home. This project combines and applies his skills in electronics, networking, programming, data analysis and much more.
From production lines to schoolyards
In just a few weeks, over 40 companies are already using this system in Romandie, the French-speaking part of Switzerland. First customers were amongst production plants. Their interests were in coupling their access verification system with a temperature sensor. Private schools were also really interested, they wanted to reassure their cosmopolitan students about their health status. But soon other sectors became interested, such as the food industry or event/congress organizers.
The CoronaSense system is simple and intuitive: the person approaches, positions their face in front of the device, and a second later they are informed of their temperature on a screen. From there, the company can add various modules such as wireless badge reading, door opening, e-mail alerts, sound notifications, a larger screen and so on.
Concentration of technologies
CoronaSense system is built around a RaspberryPi, a microcomputer the size of a credit card. It controls the various sensors (including a laser presence detector and a very precise temperature sensor) via its inputs/outputs (I/O in electronic jargon).
Specific electronic boards are designed by Adrian to connect and control other features such as speakers, amplifier, RFID/NFC badge reader, 4G SIM card or GPS sensor. The latest developments include the addition of a servo-motor that pivots so that the temperature sensor takes measurements at the ideal location on the person’s forehead. All this electronics is encapsulated in a PLA case, a beet-based vegetable thermoplastic polyester that is totally biodegradable and produced in Switzerland. The case is also designed and printed by Adrian using his 3D printer.
You now know some details about the electronic side of CoronaSense, but let’s talk about the software side, the intelligence of the system. Algorithms have been developed in order to make accurate temperature measurements and to be sure that the temperature captured is the one from a person’s forehead, and not another body part or object. You can try to put your hand in front of the sensor, it won’t be fooled. The CoronaSense system has been designed to be autonomous: it stores the data it collects, prepares its own statistical calculations for monitoring the temperature of people and sends alerts itself. It contains a server that allows the client to interact with the data through a web interface developed with the latest technologies available (client interface developed with Vue.js and a REST API made with Laravel). When a client manages several CoronaSense devices, it is possible to access a dashboard that aggregates the data from all devices:
Competitive advantages
For companies, the real interest of this system lies on three levels. First, it is extremely cheap compared to standard thermal imaging cameras worth several thousand Swiss Francs (or dollars). Second, it is extremely accurate for temperature capture (of the order of 0.1°C) compared to the infrared thermometers on the market. Indeed, it uses a high-precision 1-pixel thermal camera (also called thermopile). And finally, once installed, it is completely autonomous and will not generate any additional costs, except for a little electricity.
Come test (and approve!)
Want to see and test CoronaSense, the automatic temperature sensor? Its creator Adrian Shajkofci will be one of the main speakers at the “Our data in the face of uncertainty” event organized by the LaData association on Wednesday, September 9, 2020 in the heart of the Biopôle, in Lausanne. Full programme and registration (free) here.
