We developed a pre-prototype bracelet device for measuring a person's tiredness.
Our company has been established back in 2013 as a spin-out from Faculty of Electrical Engineering (University of Ljubljana) and we are actually quite proud of that. Also, we still have many contacts at the faculty, especially since two company owners are professors.
One of the benefits of collaborating with the faculty is the possibility to be in touch with students. We are given this great possibility to attract top-class students at the moment they are ready for either working on projects or looking for their first serious job. Another advantage of being connected to the faculty is being known for your R&D services in many laboratories. Since we have promoted ourselves many times at faculty events, we are now known for being a trusted hardware development agency that can be outsourced for enterprise-grade projects.
This time, a prominent laboratory turned to us and asked for our help. LUCAMI means Laboratory of User-adapted Communication and Ambient Intelligence and they're doing some pretty cool research in the scientific field of context-aware user interaction, and they required some special hardware for their project.
At LUCAMI, they were working on a project where vital signs of a computer user would be tracked, and the user would get informed by the computer that he needed some rest. They wanted to develop a device for detecting tiredness and attention of a person who uses a computer. The general idea was to use an accelerometer or maybe a gyroscope, put it on the person's wrist and detect movements that would indicate tiredness.
Since movements were unknown, they first needed a test group to provide feedback if they were on the right track. They set up a testing environment that contained an eye-tracking camera, a regular camera, and some Android smart watch. Then, subjects would be required to answer a series of questions, and their behavior would be analyzed afterwards.
Because the hardware that was available for LUCAMI's tests did not provide sufficient results, they decided to develop their own bracelet to replace the Android watch. They also intend to make the bracelet a commercial product after their research is finished. They wanted us to develop their own custom bracelet and embed a Bluetooth Low Energy (BLE) interface, an accelerometer, and a gyroscope. The tiredness-detection algorithm was something they wanted to provide, and we were required to include it into the bracelet's firmware. Since this was not a mass-production device, we chose an industrial enclosure and designed the hardware tu fully fit it which was acceptable for such low series.
Actually, this was not even a series but rather a prototype. Since the concept of detecting tiredness and attention using a bracelet is obviously quite new, a few other virtually trivial yet somehow important things needed to be checked before. First, if the battery drained really fast, the bracelet would heat up. The heat would, therefore, make the person feel uncomfortable wearing the bracelet. This is one of the reasons why we chose a rather big enclosure. As second, algorithms for detecting tiredness and attention were not developed yet since they first needed to record the subject's signals to develop them. And we, of course, needed the algorithms to finish the device. Quite a Catch-22. Since we had to somehow obtain the recorded data in order for LUCAMI to finish develop the algorithms based on these recordings, we speeded up a bit and rather used a Bluetooth classic interface in order to rapidly develop what turned out to be a pre-prototype that allowed recording data quite easily. Additionally, we wrote a small Python program that could be run on a PC and allowed storing the received data from the bracelet.
Quite recently, prof. Andrej Košir who is the head of LUCAMI gave a TEDx talk in Ljubljana about context-aware interactions. If you don't mind learning Slovenian (or you already know it), you are quite welcome to watch it here:
We are proud that we developed hardware that will help researchers perform state-of-the-art work and enable them developing such advanced solutions like this one. Can you imagine your computer telling you to go to bed?
Returning to the roots and developing hardware for laboratories doing advanced research work is indeed a pleasure. And no, we're not tired at all.