smarthealthcare

• This image is currently processing. Please be patient...
  

Inspiration

What it does

It helps us doing guided workout without gym trainer. It monitors the body movement and identifies the type of exercise.

How we built it

We built it using NodeMCU and MPU6050(Accelerometer and Gyroscope sensor). This two are included as a client and the sensor data is sent to the custom built server by using socket programming. Machine Learning algorithm takes this data from server using a restAPI and it identifies the type of exercise. This output is displayed on website.

Challenges we ran

Calibrating the sensors Less accuracy due to unavailability of Magnetometer sensor. Data frequency is about 2 seconds.

Accomplishments that we're proud of

wireless data transfer from nodemcu to server, machine learning for prediction of activity from the data

What we learned

Socket Programming, Machine Learning Training model, accelerometer training

What's next for smarthealthcare

Muscle movement tracking, heart beat sensing.

Built With

• arduino
• django
• machine-learning
• python
• react
• socket