SlothMotion

Inspiration

57% of Australians aged 65 and over do not meet the recommended physical activity guidelines. Physical inactivity significantly increases the risk of chronic disease, premature mortality, and disability. It is also linked to reduced bone strength, respiratory dysfunction, and a higher likelihood of cardiovascular disease and stroke. Despite these risks, many older adults face barriers to staying active. Mobility limitations, medical conditions, lack of access to exercise facilities, and discomfort in traditional gym environments can make regular exercise difficult. We were inspired to create a solution that makes physical activity more accessible, safe, and enjoyable for older Australians. Our goal is to empower people aged 65 and over to stay active by providing guided exercises they can perform from the comfort of their own home, while taking into account their individual physical limitations.

What it does

Our web app encourages older adults to stay physically active by allowing them to follow a guided exercise routine of a difficulty level of their choice personalized to their physical ailments. Our app tracks the user’s body movements and joint positions through their webcam, compares these movements with a reference routine and provides real-time feedback to help users perform exercises correctly and safely. Because our target audience may find technology intimidating, we placed a strong emphasis on accessibility and simplicity in our design. Our interface uses high-contrast colours, large readable fonts, and clear instructions.

How we built it

The frontend of our web app was developed using React with Vite, allowing us to build a responsive and simple interface tailored to older users. To detect and track body movements, we used MediaPipe Pose. MediaPipe Pose provides real-time pose estimation through the user’s webcam and extracts key body landmarks such as shoulders, elbows, hips, and knees. We implemented a pose comparison system that normalizes the user’s detected pose and compares it with a reference pose from a recorded exercise routine. By analysing differences between joint positions, the system can determine whether the movement is being performed correctly and provide immediate feedback to the user. The interface was designed with accessibility as a priority. We focused on large typography, clear navigation, and high-contrast visuals to ensure that older adults can comfortably interact with the app.

Challenges we ran into

Integrating MediaPipe Pose for real-time movement tracking was a challenge. None of our team members had prior experience with computer vision, so understanding how pose detection works and how to interpret the landmark data required a significant learning curve. We needed to learn how MediaPipe represents the human body using keypoints and how to process these coordinates in a meaningful way. Another challenge was developing a reliable method to compare the user’s live movements with a reference exercise routine. We had to experiment with normalizing pose data and defining thresholds that could detect incorrect movements without being overly strict. Designing the interface for an older audience also presented unique challenges. We had to simplify our layout, increase readability, and ensure that instructions were clear and easy to follow.

Accomplishments that we're proud of

Integrating MediaPipe pose detection and successfully using it to track a user’s body movements through their webcam was a major milestone for our team, especially given our lack of prior experience with computer vision. Another accomplishment was developing a system that compares a user’s movements with a reference exercise routine to provide immediate feedback. Our app allows users to understand whether they are performing exercises correctly in real time, helping them stay active while reducing the risk of injury. We were able to build an accessibility-focused interface. By prioritising high-contrast visuals, large readable fonts, and simple navigation, we've built an app that is approachable for older users who may be less familiar with technology. Overall, we are proud that we were able to turn an idea aimed at improving the health and independence of older Australians into a functional prototype within the limited time of a hackathon.

What we learned

We learned a lot about working with computer vision and real-time pose detection. Since none of us had prior experience with MediaPipe or pose estimation, we had to quickly understand how body landmarks are detected and how those coordinates can be used to interpret human movement. We also learned the importance of designing accessible technology. Building an application for older adults required us to rethink typical interface design and focus on simplicity, readability, and clarity. This helped us better understand how thoughtful design choices can make technology more inclusive. Finally, we learned the value of rapid prototyping and teamwork under time pressure. Collaborating across frontend and backend development, computer vision, and UI design allowed us to turn our initial idea into a working prototype within the limited timeframe of the hackathon.

What's next for SlothMotion

We plan to expand SlothMotion to better support a wider range of physical restrictions and medical conditions, allowing more users to safely participate in exercise routines tailored to their needs. We also aim to improve accessibility by integrating voice controls, enabling users to navigate the app and start routines without needing to interact heavily with the interface. Another feature we could implement in a future iteration is an AI assistant that can guide users through the app and answer questions. We also plan to introduce daily reminders and notifications to encourage users to maintain consistent exercise habits.