Peloton Guide Doesn’t Correct Your Form. I Think I Know Why

Peloton has announced their plan for launching their latest product: Peloton Guide will debut April 5, 2022 and there is one element about this event that stands out: its peculiar announcement has me intrigued. Priced at $295 for just the Guide on its own, this tool enables you to observe your posture directly on screen during strength classes with an instructor. No additional equipment such as Peloton Bike or Tread is necessary – simply a TV will do. Though initially focused on strength classes, my prediction is that they’ll gradually expand its use to other class categories – like yoga. A recent public release of an unplanned yoga class included an instructor discussing her use of the Guide and providing insight into their testing efforts.

My confusion stems from their promotional materials: there is no mention of form corrections or feedback services in any way.

By tuning in, you are able to view yourself onscreen and compare your actions with those of an instructor. Although it vaguely monitors your movements, this does not provide specific guidance on necessary adjustments – instead acting more as a virtual instructor than anything else. Furthermore, Peloton acquired Otari Studios around December 2020; these specialists had developed software specifically tailored for yoga mat workouts which offered AI-powered feedback through Otari’s AI product, Otari Feedback; so one might assume Peloton already possess this technology?

These factors, in my estimation, provide strong support for the absence of form pointers in the Peloton Guide. Naturally, these are only my personal speculations but seem likely.

Why the Peloton Guide Does Correct Your Form…Yet? My Personal Opinion

Otari Didn’t Have The Technology Built

Otari’s Indigogo
via Otari’s Indigogo

Peloton may have purchased Otari more with an eye to acquiring patents than on purchasing fully developed technology, as Otari hadn’t reached maturity in its development stage yet and therefore Peloton decided on starting from scratch with Otari.

Although this might have been an unfortunate circumstance for Peloton, I believe they would have had ample time to improve their technology since then. With some of the finest product and engineering teams at their disposal, I believe they would have developed user-satisfying solutions at this point – particularly given that many current products already provide exercise identification capabilities as well as form recommendations.

Peloton Has The Technology, But Not The Data

Peloton may possess both the technology infrastructure and dataset necessary for effective AI operation, yet lacks an adequate data intake process in machine learning – this data would include workout videos, categorizing of exercises and benchmarks for accurate form.

Consider Tempo, a connected fitness product focusing on home-based strength training that integrates form correction. This innovation was born after Pivot (then known as Pivot) had collected ample data for use by their AI system. Prior to developing Tempo, SmartSpot had been utilized by trainers in gyms allowing Tempo to amass over one million labeled workout sessions which ultimately proved instrumental in training their new AI model to provide form feedback feedback.

Peloton’s release of their Guide opens up an opportunity to gather data more efficiently. By categorizing workout releases into categories and noting when specific exercises start and finish, Peloton could more efficiently collect their information – an approach made clear in their promotional material which showcases movement tracking mechanisms in action.

Otari’s Indigogo

Imagine Ben Alldis engaging in tricep kickbacks between 1:25 to 1:55 of a video. Peloton could analyze a Guide user’s video during that workout segment and classify it automatically as a demonstration. Peloton’s advertised movement tracking functionality would aid them in selecting those videos suitable for machine learning processing and eliminating potentially non-useful clips; this automated approach offers several advantages over manual categorization as it provides richer real-world data which could not otherwise be produced internally or acquired externally by other sources.

It’s a Business Decision

At Peloton, it seems likely they’ll integrate form feedback into their system eventually, and may have already laid out an implementation strategy or roadmap to do so gradually. With their penchant for staged releases and gradual upgrades, I anticipate seeing new functionalities added gradually over time–perhaps by September–when businesses traditionally unveil holiday season offerings–we hear of an announcement detailing its implementation within Peloton Guide.

Read also:


Leave a Reply

Your email address will not be published. Required fields are marked *