Bettering bicycle security with voice-activated flip indicators

Bicycle security is at all times a priority, however is a very massive downside in cities the place biking is much less frequent. Drivers in vehicles don’t count on to come across cyclists and don’t know learn how to share the highway correctly. To guard themselves, cyclists should be as predictable as doable and overt with their intentions. Signaling is crucial, however hand indicators could also be misinterpreted. That’s why Manivannan developed an AI-enabled good helmet that leverages Edge Impulse and Arduino for the hands-free voice activation of flip indicators.

Edge Impulse makes it simple to coach machine studying fashions and deploy them on the sting with embedded units. On this case, that system is an Arduino Portenta H7, which has a strong STM32H7 microcontroller and was designed particularly for purposes that depend on laptop imaginative and prescient, machine studying, and different resource-intensive duties. 

To coach the Edge Impulse mannequin, Manivannan solely wanted three audio units: one of many phrase “left” for the left flip sign, one of many phrase “proper” for the best flip sign, and one with examples of ambient background highway noise. Street noise samples had been essential, as a result of they inform the mannequin what sound isn’t a voice command and is secure to disregard.

The {hardware} is easy, with energy coming from a 5V battery pack. The flip indicators are strips of LED “filament” managed by the Portenta H7 in response to the detected voice instructions. The Arduino resides within a plastic container for climate safety and that, together with the LED strips and battery pack, strap on to a store-bought bicycle helmet.

Now Manivannan can cycle safely, signaling to others on the highway with out taking his fingers off of the handlebars.