Inspiration

Fitness is an essential part of well-being, but lack of guidance when working out can lead to injury, and deter people from building healthy habits. One in five people get injuries from bad form when exercising, and bad form can also stop you from progressing To tackle this issue, we built an AR fitness training app that reduces the risk of injury during workouts and improves efficiency by providing users with a visual model of their form, along with specific feedback and technique corrections, in augmented reality.

What it does

MoveU displays a life-sized, personalized avatar that mirrors you when you’re exercising – no camera setup is needed. Our solution allows you to see your full body motion during a workout, along with specific guidance depending on the user’s workout form. This type of feedback normally requires an expensive personal trainer, but with MoveU, you see exactly how you can correct your form and can be confident that you’re making the most of your workout.

How we built it

We built a prototype using Unity and the Sony Mocopi motion sensors, and tested our solution on Quest 3. We also leveraged ShapesXR for design and prototyping that includes future goals for MoveU.

Challenges we ran into

The Sony mocopi sensors are relatively new, so our tracking solution isn’t as precise. We account for this in Unity by adjusting our exercise technique thresholds to better accommodate the sensors.

Accomplishments that we're proud of

We’re new to AR development with Unity, so it was an exciting challenge to learn Unity development over the weekend.

What we learned

We learned a lot about the challenges of body tracking and pose estimation. After exploring different solutions, we decided to use the Sony mocopi sensors as it would enable us to build an experience that requires less setup from the user. Rather than trying to record yourself at the gym or use an entire external camera setup, you can simply start exercising and our AR experience will guide you to excellence.

What's next for MoveU

We will launch our current fitness demo as an app that has improved sensor data processing and calibration to track precise movements from users. We’re aiming to provide a comprehensive exercise experience that can provide a similar level of quality to personal trainers, but at an affordable price.

Built With

  • arkit
  • c#
  • coreml
  • createml
  • quest2
  • quest3
  • shapesxr
  • unity
Share this project:

Updates