Cure.fit acquires AI body tracker Onyx to enable two-way workout videos

Selling prerecorded exercise videos for at-home workouts has been a viable business model for decades, and the pandemic has only increased the need for home solutions spanning every conceivable fitness category. But these videos couldn’t replace a key element of in-person training — the individual attention from a personal or group trainer. Now India’s best-funded health startup is making a move to bridge that gap, using a California company’s AI system to replace human fitness coaches in users’ homes.

Currently available as an iOS and Android app, with support for Apple TV and Chromecast display devices, the Cure.fit platform offers both exercise videos and online group classes, falling back to a basic “energy meter” as a simple metric of a user’s activity while working out. Cure.fit’s acquisition of Onyx, a company with an eponymous fitness app, will enable it to replace the “energy meter” with a sophisticated computer vision system that can segment a user from their background, recognize various types of workout activities, and automatically count reps without need for a human coach. Onyx can also understand a user’s performance, correcting issues with form and offering properly timed motivations, just like a human coach.

Cure.fit’s acquisition of Onyx is significant for technical decision makers because it demonstrates how applied computer vision AI technologies will upend or dramatically improve well-established businesses. As impressively produced as Cure.fit’s workout videos may be, their limited interactivity is a disadvantage compared to new competitors such as Apple’s new Fitness+ app, which can use wearables to track biometric information. Computer vision and AI will enable Onyx and Cure.fit to track motion at scale without the need for wearables, freeing their users from needing to visit gyms, while giving trainers — or tech companies — the ability to actually coach larger numbers of people without physically interacting with them.

Onyx describes its technology as “the world’s first 3D motion capture system on your phone” and says it’s capable of tracking “nearly any exercise,” from squats to planks, push-ups, kicks, and running and jumping in place. The user places a smartphone in a location capable of seeing the full body during exercise, and the app uses computer vision to render the user as a silhouette while automatically comparing limb and torso motions against ideal representations. It’s an impressive example of interpreting and translating human motions into live, actionable data.

Adding Onyx’s real-time analysis will help Cure.fit improve the personalized measurement element of its platform. Cure.fit will be able to expand the range of tracked activities to include more nuanced exercises, such as dance and yoga, as well as fast-paced activities like high-intensity interval training, while accurately gauging a user’s compliance with proper forms and pacing.

The company already offers an ambitiously broad collection of health-related services, ranging from exercise and diet assistance to medical care, online mental therapy, and guided meditations, though the services tend to be either one-on-one or depersonalized for mass distribution. Today, users can participate in group classes either at select physical locations or via video calls at home, but scheduling classes with the company’s human trainers could prove challenging; the company’s website already flags potential class-related time zone issues for users outside India.

Backed by over $400 million in funding, Cure.fit expects Onyx to help it build a “hardware agnostic AI-led platform” that uses technology to provide humanlike personalized guidance across multiple dimensions of health, ranging from physical and mental wellness to nutrition. The highly rated Onyx standalone app remains available for download in its prior form, offering $15 monthly and $100 annual subscriptions. It’s unclear when the standalone app will be folded into Cure.fit’s lineup or how its pricing will change as a result.