These clever slippers sense common actions and falls utilizing machine studying

With regards to exercise displays akin to smartwatches, rings, and pendants, they’re typically thought of cumbersome or too tough to maintain monitor of, particularly for the aged with reminiscence or dexterity issues. For this reason the crew of Jure Špeh, Jan Adamic, Luka Mali, and Blaz Ardaljon Mataln Smehov determined to create the SmartSlippers undertaking, which is a much more built-in technique for detecting steps and falls.

The {hardware} portion of the SmartSlippers prototype is only a Nano 33 BLE Sense board on account of its onboard inertial measurement unit (IMU) and Bluetooth® Low Power functionality. At first, the crew collected 14 minutes of 5 various kinds of actions: strolling, operating, stairs, falling, and idle throughout the Edge Impulse Studio. From right here, they skilled a neural community on these samples, which resulted in an accuracy of round 84%.

With the Nano now in a position to detect movement, the following step was to get the board to speak with a cellphone in order that emergency providers could possibly be known as within the occasion of a fall. Their firmware units up a BLE gadget and provides a attribute that sends information to the related cellphone when an occasion happens. And eventually, their customized Android cell app shows the present standing of the SmartSlippers and may even name somebody if a fall is detected.

To learn extra in regards to the undertaking, you may go to its write-up right here on Hackster.io.