Social Distancing using Bluetooth

Inspiration

Social distancing has become a popular method being adopted to flatten the COVID19 curve. Humans are extremely social species which makes social distancing an effort of high discipline.

What it does

Each device keeps scanning Bluetooth devices around it. When a device is in range, the signal strength is obtained from the broadcasted signal. The algorithm uses the signal strength to estimate a distance value. If the distance is less than 6ft or 2m, the user is alerted.

How I built it

I started the effort by trying to estimate the distance between two Bluetooth devices. What I concluded in the online research that in a closed environment the RSSI(log scale) vs distance curve is almost a linear curve. Here, I believe the same model can be used to implement social distancing. I have run the code on my laptop. The Bluetooth of the laptop scans the Bluetooth devices (mobile phones, fitness trackers, etc.) and fetches the RSSI value of each device. Once the value is below the permissible limit, there is a notification alert for the user.

Challenges I ran into

RSSI values often fluctuate, which may cause a false alarm or sometimes missed detection. To counter this the algorithm takes repeated values of the RSSI and calculates the mean and variance of the values. The values are fed into a probabilistic model for accurate detection.

Accomplishments that I'm proud of

The idea of using such a simple and subtle technology such as Bluetooth can help fight COVID19, feels like an accomplishment.

What I learned

To reduce errors in detection by making a probabilistic model.

What's next for Social Distancing using Bluetooth

To implement it on an Andriod/ iOS device. To improve the distance prediction.

Built With

• bluepy
• python