Improving Cyclist Safety with a Smart Glove Signaling System
This wearable device uses an Arduino running TinyML to send hand gestures to a backpack-mounted LED matrix over BLE.
Matlek rides over an hour each day to work in Marseille, which is reportedly the worst city for cyclists in France. In an effort to help road users communicate more effectively and enhance safety, he’s come up with an innovative smart glove and backpack signaling system.
The "glove," or hand-mounted controller, features an Arduino Nano 33 BLE Sense, battery, and little else. It uses the board's built-in IMU to detect hand movements and runs a TinyML gesture recognition model to translate them into rider intent. This data is passed along via BLE, including left and right twists of the wrist for turn indicators, a back twist for stop, and a forward twist as a scrolling greeting.
The backpack-mounted LED panel is driven by an ESP32 board, receiving BLE signals from the hand unit. A level shifter is also implemented to enable the ESP32’s 3.3V logic to properly work with the nominally 5V 20x8 WS2812B display.
The LED matrix reveals turn/stop/greeting signals according to the BLE interface, as well as a default green up arrow to show that that Maltek is riding in a nominally straight line.