Inspiration

  • 23.8% of Singaporeans will be aged 65 and above by 2030
  • More than 1 in 4 elderly experiences a fall annually
  • Falls are especially dangerous for the elderly
  • Timely assistance could prevent the exacerbation of injuries
  • It would be useful to see who is more prone to falls

What it does

  • SafeSteps is a wearable IoT device that can be strapped to the arm
  • Detects when the wearer falls
  • Immediately sends a Telegram message to family members, alerting them
  • Also displays falls in a dashboard to see if the elderly are very prone to falling, thus requiring more care
  • Allows doctors/nurses to also monitor all patients under their care

How we built it

  • We used edge computing to perform computation on a low-performant device using embedded-C.
  • We used an Arduino board and used Tensorflow-lite to deploy a Machine Learning model in it.
  • Fall detection is performed within the Arduino itself.
  • We wrote python scripts for an RPi server. Arduino sends data to RPi through serial communication.
  • RPi server is connected to the internet and updates the Firebase real-time db.
  • RPi also calls the Telegram API to send a text to the family member.
  • Family members/doctors can monitor falls through a web client.
  • They can see past history of falls to see if the elderly are more prone to falls, thus getting more attention.
  • Web client was built using React, Typescript, Tailwindcss.
  • When a fall occurs, the client gets a push notification as well.

Challenges we ran into

  • There were a lot of problems with the Arduino communication to the server and integrating the components into a single functional service.

Accomplishments that we're proud of

  • Working together as a team and under pressure to build a successful project that can be used by the public to monitor their safety.

What we learned

  • Hardware development
  • Machine Learning in edge computing devices
  • Front End Web Development
  • Server Side Development

What's next for SafeSteps

  • Bring this awesome product to the market! 🚀

Built With

Share this project:

Updates