Inspiration

“Cyan is a restful, calming color that symbolizes relaxation, tropical locations and vacation” describes Wikipedia. Adults spend up to 11hrs a day looking at the digital screen and due to the COVID-19 pandemic, there has been an excessive and unavoidable increase in screen time. Our inspiration was our own struggle to stay healthy while being productive as an employee or a student. This led to the inception of a simple and lightweight chrome browser extension called Cyan which contextually reminds users at intervals to correct their posture and take well deserved breaks in order to get into the habit of having a healthy life while using their computer.

What it does

Cyan is an inclusive and health-conscious, chrome browser extension targeted at users who spend a maximum amount of time in front of their computer screens. It leverages the power of machine learning to monitor a user's posture in real-time while also providing contextual feedback and corrective measures. It also aids users in regulating their work and break cycle by reminding users to take breaks and provides optional content to relax. The feedback and notification would be implemented with both visual and auditory cues to aid a diverse range of users.

How we built it

We used the Google teachable machine platform to train our posture model and hosted it on google cloud. We trained our machine learning model with four classes/categories like Good Posture, Bad Posture, Too close to screen and Using other devices (for example: being distracted by mobile devices). We then used Figma to prototype the user interface for our final submission

For our process, we used the design thinking approach to find a balance between users, technology and business goals. See more details (notes, maps, user flow) please visit team's Google Doc. In doing so,

  1. Interviewed 2 users to understand their screen time behavior,
  2. Weighed in the pros and cons of the currently available solutions for posture correction and relaxation ( UPRIGHT GO, Calm, Productivity pet, and others)
  3. We identified the interface/interaction types, must have and nice to have features
  4. Identified different categories to train the ML model
  5. Created user flows, sketches, wireframes and then the final prototype

Accomplishments that we're proud of

As first time hackers who had never before used machine learning we hoped to use this exciting technology to aid healthy living while also ensuring minimal intrusion. We successfully implemented an interactive prototype that combined the power of ML, teachable interfaces, user-driven insights and accessibility within 24hrs which would surely benefit users in their day-to-day routine.

What we learned

We learned so much more about the applications of emerging technologies like computer vision, machine learning especially in the healthcare domain which we had never before ventured into. A very important lesson learned during the process of training the model was the presence of bias and how we need to be cautious and mitigate biases for safer use. We were exposed to different design principles and guidelines for browser extensions as well as the need to make it accessible to all.

What's next for Cyan

The next steps would be to implement the prototype in the production environment with browser friendly and secure protocols and programming languages. It would be great if companies would adopt this approach to encourage their employees to live healthy lives. The ML model will have to be trained with a larger and diverse dataset in order to identify correct postures while sitting and standing and be more inclusive.

Built With

Share this project:

Updates