Headaches, jaw pain, TMJ, and tooth damage are things that all of us would like to avoid, but for those with jaw clenching issues (or bruxism), they can be regular occurrences. But a hardware hacker named Andrea would like to make these problems a thing of the past. The solution Andrea came up with is a device that trains users to stop clenching their jaw, and since this problem most often occurs at night, it works even when the user is sleeping.
To support a wide range of use cases, the system can detect jaw clenching, log these occurrences, and provide alerts when clenching events are detected. In this way, the detector can take a backseat approach in which it simply tracks events for the user to evaluate at a later time. But when a definite problem has been identified, it can also take a more proactive role. In that case, the system can trigger alerts that get the user’s attention — day or night — to interrupt clenching and allow for behavior modification.
A report generated by the system (📷: Andrea)
Another interesting feature of the device is that the logged data can be turned into reports that correlate clenching with other events. This is made possible by tagging key events during the day. This would reveal, for instance, if particular foods or activities tend to be correlated with an uptick in clenching events.
The device is built around an Arduino UNO R4 WiFi development board and an Olimex SHIELD-EKG-EMG that measures muscle activity in the jaw. This hardware is housed in a 3D-printed enclosure that can either be mounted on the wall for nighttime use, or clipped onto a belt for daytime use. A battery module keeps it powered up on the go. The hardware design and software are all open source, so anyone can make their own copy of the unit. It is also affordable, with the parts coming in at around $120.
It might not be the most popular option for mobile use, however. Electromyography signals are captured via a set of three electrodes that are attached to the forehead with a conductive gel. As such, this system may not be popular with the casual biohacker, but for someone with major bruxism issues, it could be a big help.
The companion app helps to associate life events with clenching (📷: Andrea)
Data from the sensors is analyzed with a machine learning algorithm (a Support Vector Machine) to detect jaw clenching events. This information can be viewed in a companion app built for Android phones.
The system is not quite ready to go after assembling the hardware. The machine learning model needs to be trained before the first use. The sensor also needs to be calibrated. There is a large amount of documentation to help you get started, however. So if you have the interest, and a little bit of patience, you should not have too much trouble. Check out the project write-up for all the necessary details.