ROMWOD is an acclaimed fitness app boasting a vast library of mobility training videos. Subscribers are offered daily workouts to maximize athletic performance, resolve pain and reduce risk of injury.
After designing a new app ecosystem, I created a feature to help people test their mobility and start a tailored training program, all from the comfort of their home.
I designed a 2-position, 3-angle mobility test which captures essential range of motion and watches for potential dysfunctions across the body.
To streamline the experience and maximize the accuracy of recorded data, I created a vocal and visual interface which uses real-time body and environmental hints as a personable way to coach the user through.
The full-screen teaser provides quick access to mobility test, while a new menu item allows manual navigation
to the new feature.
The first onboarding screens help users get familiar with the mobility test, collect valuable well-being data and offer two ways to take the test, depending on whether the user is on their own or doing it with a friend. This selection informs which camera (front or back) is preselected for the test.
After initially experimenting with ARKit, Apple's Vision framework revealed itself as more accurate despite being limited to 2D, with the added benefit of providing significant performance gains.
I defined an array of metrics to form a reliable picture of a user's mobility and we used Vision's native body joints as coordinates to measure the corresponding angles. These were then compared to what is consider healthy range of motion as a foundation for our scoring system.
Moving beyond mere range of motion, I selected a number of common joint dysfunctions which we were able to detect, quantify and grade using the same approach.
The "Get Ready" screen, last before the test begins, provides crucial tips to optimize results, as well as
offering optional settings for recurring users who wish to further customize their test experience.
Visual design patterns evolve as the test progresses, providing real-time feedback that the user's body is
recognized and being analyzed. Audio and visual cues make the experience not only accessible but absolutely
immersive. As rich and interactive as it is, the mobility test takes under 2 minutes to complete, after which an animated
loader augmented by sound effects builds anticipation in advance of discovering test results.
Tapping on the overview scorecard opens the detailed view, where users can access position scores, individual joint scores, critical joint, and detected issues. Each issue can be tapped for even more insight into a user's Mobility.
After extensive internal testing, ROMWOD Mobility was initially released as a beta feature to our power users, pending further developments before giving access to a wider audience.
We built the entire experience with a large panel of tracked events linked to Amplitude, allowing us to swiftly zero in on pain points and identify which sequence of events led to a potential crash or scoring inconsistency.
I designed a survey to gather critical feedback from users, linked to Airtable for data analysis. The feature was exceptionally well received by users and our roadmap revealed to be ideally aligned with their needs, priming ROMWOD Mobility for public release.
The feed of recommended routines allows users to quickly jump into a routine and improve their Mobility. A dedicated content card makes this feed accessible from the Explore tab as well. Content preferences can be adjusted to fine-tune recommendations and filter out unwanted videos.
The scorecard on the results overview can be swiped, giving access to a user's Mobility stats card. This summary of stats provides an indication of progress, the most critical joint, and the most improved joint over the last six tests. Tapping the stats card opens up the full stats view, where users can select two dates to draw stats between and start diving through their data.
The history feature makes it easy to review the frequency of a user's tests and swiftly jump back into a past assessment to compare scores and data.
The survey helped us troubleshoot user issues and gauge their desire for new features, while the debug mode allowed us to collect their biometric data, videos, and snapshots for a more in-depth analysis while complying with personal data concerns and regulations.
ROMWOD Mobility was gradually made available to an increasing number of users and devices, as we collected more data and used it to further improve the feature.
We welcomed a physiotherapy specialist to the team, who brought a holistic view of mobility to our scoring system. Armed with this new approach, we repurposed our method to focus on joint function, both individual and combined, and provide users with a deeper understanding of their movement quality.