Cough Activated Face Visor

A HCI concept demo about cough detecting visor

Cough Activated Dynamic Face Visor

A prototype developed during the COVID-19 pandemic to explore advanced personal protective equipment (PPE) through wearable technology. The pandemic rapidly affected everyday life, creating challenges and instigating the development of various solutions aimed at preventing disease spread or making early diagnoses.

My role in this project involved the technical implementation, programming, and design of the device, bringing the conceptual framework to a functional reality.

Concept & Functionality

The core of this dynamic face visor is its automatic response to coughing sounds. When coughing is detected, the transparent visor screen promptly moves to cover the wearer’s face, whether the cough originates from the wearer or those nearby. The visor automatically resets to its open position after a defined period without further coughing, for example, one minute. This operation provides a tangible visual cue of potential risk, subtly encouraging both the wearer and observers to adopt better hygiene practices and maintain social distancing.

Technical Implementation

The prototype was built by modifying a standard 3M face visor. Its smart functionality is powered by an integration of hardware and software components:

• Microcontroller: An Arduino Nano 33 BLE Sense serves as the central processing unit, chosen for its on-board microphone which was used for capturing audio.

• Actuators: The mechanical movement of the visor is achieved using two servo motors.

• Structure: Custom mounting hardware for the servo motors was 3D printed, integrating components with the visor frame.

• Cough Detection: The system leverages a TinyML machine learning model for real-time cough detection, based on an implementation developed by the Edge Impulse team as part of the UN Covid Detect and Protect Challenge.

The model utilized an initial dataset provided by Edge Impulse, which enables it to differentiate only between quiet background noise and a small range of coughs.

• Programming: The system’s operation utilized additional Arduino code based on resources provided by Circuit Digest.

Project Goals & Observations

Beyond direct protection, a primary aim of this demo was to spark discussion and ideation on smart PPE that activates only when necessary

, making it both more effective and less intrusive to wear. Initial ad hoc trials indicated that the dynamic visor provides a tangible indication of possible risk to the wearer and people in their vicinity. Through this mechanism, the prototype can potentially influence behavior, for example, encouraging correct coughing etiquette by covering the mouth. We also observed a notable level of surprise in the wearer when the visor automatically activates.

Future Outlook

While the prototype presents an explorative concept, future development areas include exploring the possibility to design such transformations in a less disruptive manner, improving the performance of the cough recognition model by addressing the limitations of the initial training dataset, and investigating cloud-based analysis solutions and the integration of motion sensor data.