Physio24, a thriving, forward-looking physical therapy practice, wanted to create an app that would:
They wanted to explore innovative technologies like AR, voice UI, and motion capture, in features like a gamified PT program, tracking dashboards, and self-reporting.
I completed this project as a student in Hyper Island’s UX Upskill program.
I started by mapping out the as-is user journey for a provided persona, Anara. Anara represented a tech worker in her mid-20s, who had gotten a sport injury and was seeking physical therapy she could fit into her busy schedule.
I based the stages of service on my past experiences with physical therapy, remembering my high and low points and reflecting on how things would be similar or different for Anara, and giving extra attention to the areas that overlapped. The user journey map would show that I could address many of the business goals and user pain points by working with the at-home therapy session. What would it look like, and what role would advanced technologies play?
Focus on 1-on-1 activities (manual adjustments, teaching new techniques) or special equipment use, during intermediate sessions, and let users do the rest at home.
Help patients create a habit around physical therapy at home and remind them of sessions
Track space and equipment required for each technique and match it to what the user has available. Allow user to filter or select alternatives
List or library of assigned exercises and treatments with instructions, video, and tools to reinforce correct technique
I then did some desk research. I looked into physical therapy in general: how could plans be adapted to fit into Anara's busy schedule? What makes a successful PT session? How do therapists measure progress?
I also analyzed some of the competition. I looked at apps across the fitness, health, and wellness spaces, looking at features, design patterns, level of interactivity, and approaches to XR. I also completed workouts in two apps and analyzed the successes and failures of each.
Having analyzed the as-is experience, now I began to explore how it could be improved through technology. Some kind of AI coach? A fully voice-driven experience based on flawless natural language processing? To explore this, I created an extremely blue-sky task flow, then crafted it into a narrative dialog between Anara and the app.
This process helped me to:
Based on the research & narrative exercise, I began designing, eventually prototyping an app that would:
Depending on budget and timelines, this solution could be implemented using AI for natural language processing, or using more traditional methods.
I started with a higher-level task flow, as shown below, to explore how to integrate completing exercises with wellness and reflection. I expanded that to a more complete user flow, iterating as I sketched and designed screens.
A dashboard tracks streaks to build habits, shows scheduled sessions, and provides a few alternate treatment options, usually quicker sessions with different focuses.
Session contents will automatically adjust if a user is missing equipment or lacking space to do certain techniques.
During a session, users can complete their treatment using their camera as a live mirror or with a looping video. They can choose to use voice commands and voiceover or to work silently.
The app collectes feedback after each exercise. This both provides valuable data to the PT and app and encourages the user to reflect and pay attention to how their body feels during each step.
At the end of a program, users are reminded of their upcoming sessions. At-home and in-person sessions are given equal weight, emphasizing that at-home treatments are just as important.
Interactive prototype of full user flow, from notification through completing a program and booking the next session.
Allowing adaptations to schedule and context of use, like equipment and space, and simply helping users to keep track of which exercises and how to do them, can solve pain points in both the remote and onsite physical therapist relationship.
Wellness content about subjects like breathwork is cheap to produce, and quick to do for the end-user — and studies show real physical and mental health benefits. Providing that content in a quality platform where the user has already established habits will be a great way to keep users coming back.
I didn't start this experience expecting to come up with a VUI solution — I generally find voice-based systems incredibly annoying. But when your phone is propped up on a chair, voice is a game changer. I realized that I was thinking of VUI as Alexa, when really, it could be more like Yoga with Adriene, or your favorite fitness teacher, providing useful insights and reminders of technique when you really needed them.
I really enjoyed using the Fyter app, but in the days afterward, it was very clear I'd pushed a bit too hard. Focusing on kicking marks on screen as quickly as possible came at the expense of paying attention to how I was moving. AR fitness apps should be designed carefully, encouraging periods of reflection or helping to achieve the correct form as much as encouraging users to meet gamified goals.
I was concerned that a mixed-reality solution would be complicated under real-world constraints, or confusing to set up. Based on the apps I tested, that doesn't have to be true. A chair and a water bottle are all that's needed to prop up a phone, and it seems that, if a user can make space to do an exercise video, they can likely make space for AR.
I was one of 3 students selected to present my project to the real client who inspired this brief, a physical therapist currently seeking funding in early 2025. We're awaiting the outcome of that fundraising to proceed.
If I were to proceed with this project, my next step would be to do some basic user testing with the ProtoPie prototype, updated with physio-designed content for a real 5-minute program.
I'm also very curious to explore the experience from the physical therapist's side. What data, and how much, would the physical therapist need to receive from the client? How should it be presented? Are there opportunities to use post-processing to show overlays or get metrics from user-submitted videos? I'm always interested in improving tools used for work, so there are dozens of interesting questions there, as well.
