Adaptive Fitness Technology: What Works Now, and What the AI Boom Is Missing
- Shane Riddle
- 2 days ago
- 12 min read
Updated: 21 hours ago
In England, disabled people are still about twice as likely to be inactive as non-disabled people. In Australia, around 14% of adults with disability take part in sport at least weekly.
Those gaps have proved stubborn, and for three years the fitness industry has promised that artificial intelligence and personalised coaching would get more people moving. In this post we'll look at how much of that promise currently reaches adaptive athletes, why the technology struggles with bodies that move differently, and, most usefully, which adaptive fitness technology genuinely works today. If you want the wider picture of where this is heading, our post to wearables as coaches covers the shift from recording to recommending.
In June 2026, one of the sector's most-read trade titles published its annual technology outlook with 66 pages and 20 articles on where fitness technology goes next. None of the 20 covered disability, accessibility or adaptive sport.
Key Takeaways:
Your tracker may be guessing: A 2024 peer-reviewed study found that neither the Apple Watch Series 4 nor the Fitbit Versa was accurate enough to estimate energy expenditure or heart rate during wheelchair propulsion.
The participation gap is real and measurable: In England, 39.3% of disabled people are inactive compared with 19.7% of non-disabled people, a gap of 19.6 percentage points.
Real adaptive fitness technology exists now: Wheelchair-accessible strength equipment, push-based activity tracking and a recognised inclusive fitness qualification are all commercially available today.
The trade agenda hasn't caught up: One of the sector's most-read annual technology outlooks ran to 66 pages and 20 articles in June 2026 without covering disability, accessibility or adaptive sport.
Human coaching still wins, everywhere: Only 10% of more than 10,000 consumers surveyed globally prefer AI workout guidance to a human coach, and the youngest age groups are the most resistant.
You have real places to start: Disability Sports Australia, Move United, Activity Alliance and WheelPower all run or list programs you can contact directly.
Table of Contents
How Big the Gap Actually Is
In England, Sport England's Active Lives 2024-25 data, as analysed by the charity Activity Alliance, found 39.3% of disabled people were inactive compared with 19.7% of non-disabled people. That is a gap of 19.6 percentage points, and it widened slightly on the previous year. There is some progress with 49.1% of disabled people now meeting the 150-minutes-a-week threshold, the highest figure recorded since the survey began. But disabled people remain about twice as likely to be inactive.
In Australia, the Australian Sports Commission's Clearinghouse for Sport reports AusPlay figures showing that in 2024-25 an estimated 492,000 people with disability aged 18 and over (14%) took part in sport-related activities at least weekly, split 17% of men and 11% of women. Around 1 in 5 Australians, roughly 5.5 million people, have disability.
Those are large populations with documented, persistent barriers to activity. They are also, in commercial terms, underserved. Technology that closed even part of that gap would matter enormously.
Where Personalisation Actually Breaks Down
The calorie figure on your watch is not a measurement. It is an estimate, with the device reading motion and pulse and then applying a proprietary model to turn those signals into a number. Because the model is a trade secret, when the number is wrong nobody outside the company can say precisely why. What can be tested is whether it is wrong at all.
In 2024, researchers tested it with a cross-sectional study published in JMIR Formative Research by Danielsson, Vergeer, Plasqui and Baumgart. They put 40 participants, 20 manual wheelchair users and 20 people without disability, through wheelchair propulsion on a treadmill at varying speeds and inclines across three test days, measured against indirect calorimetry and a chest-strap heart rate monitor. The Apple Watch was set to its wheelchair-specific mode, Outdoor Push Walking Pace. The Fitbit has no wheelchair setting, so it was run in treadmill running mode.
The Apple Watch Series 4 underestimated energy expenditure by 27.4% in wheelchair users and 32.1% in people without disability (mean absolute percentage error). The Fitbit Versa overestimated it by 73.9% and 44.7% respectively. Heart rate held up better, though both watches read low: the Apple Watch was out by 8.5% and 8.1%, the Fitbit by 17.4% and 14.3%.
"Neither the wrist-worn AW nor Fitbit were sufficiently accurate for estimating EE or HR during wheelchair propulsion." — Danielsson et al., JMIR Formative Research, 2024
The authors cautioned against relying on these devices to regulate training intensity or assess energy balance in wheelchair users.
Those paired figures aren't what I expected. The Apple Watch was not worse for wheelchair users. It was marginally worse for the control group doing the same activity. Across almost every comparison in the study, the errors didn't differ significantly between the two groups. The only exception was heart rate on the Apple Watch, which read about 5 beats per minute lower in wheelchair users.
The twenty participants who do not use a wheelchair got numbers just as wrong as the twenty who do. Whatever the watch is failing to read, it is not the disabled body. It is the pushing.
The Apple Watch never asks a user about impairment type or level, and the authors note its results suggest it can't infer those characteristics from movement or heart rate either. Comparing their figures with a 2020 study of the Series 1, they found no meaningful improvement in the wheelchair-specific algorithm across three generations of hardware. Apple built the mode, then left its accuracy where it was.
The one place the study does find a population effect is sex. Estimates from the Apple Watch were significantly worse for the women, and the authors connect that to earlier work indicating wearable algorithms are built largely from reference data collected on men.
The researchers can't say why the watch reads low, because the algorithm is closed. Apple may be reporting conservative figures on purpose, hedging against people eating back exercise they haven't done, or the development data may have come from wheelchair users with lower fitness or higher levels of spinal cord injury than the group tested here.
If a calorie figure can be out by a third in one direction or three quarters in the other, every feature built on top of it inherits the error, the recovery score, the readiness ring, the "you should take a rest day" nudge, the AI coach's next recommendation. Nothing on the screen tells you the number is unreliable. This is also why questions about who owns and interprets athlete body data land differently for adaptive athletes than for anyone else.
The authors point to two levers. The first is better sensor hardware, which currently struggles to hold a clean heart rate signal during activity at all. The second is better algorithms, which could improve without any new hardware by taking account of personal characteristics such as training status, impairment type and level.
What the Industry Isn't Building
A reasonable proxy for what the fitness sector plans to build next is what its trade press chooses to cover. Athletech News, one of the industry's most-read titles, published its fourth annual Technology & Innovation Outlook in June 2026. It runs to 66 pages and 20 articles covering AI on the gym floor, wearables funding, virtual reality, body composition scanning and the economics of personal training. None of the 20 covers disability, accessibility or adaptive sport.
One report is one report, and it would be unfair to read a whole industry from a single contents list. But it's a fair indicator of which conversations are currently loud, and this one is not among them. Interestingly, a disabled athlete does appear in the report's photography, but not in any of the 20 articles.
Adaptive Fitness Technology Available Now
The picture is far better on the ground than the trade agenda suggests. All of what follows is commercially available and in use today.
Strength equipment is the clearest case. HUR's accessible line uses pneumatic air resistance with a zero starting load, and the machines are built so a wheelchair user can roll directly into position; a rollaway seat means the same machine still serves everyone else. Speedflex took a different route with the Blade, a seated machine whose resistance adjusts separately for the upward and downward phases of a movement, which matters for people with reduced arm strength. It went into Gym Possible, a charity gym in North East England, in February 2026.
On the consumer side, the wheelchair mode is already sitting in one of the world's most widely worn wearables. Once it is enabled in the Health settings, the Apple Watch counts pushes rather than steps and offers Outdoor Push Walking Pace and Outdoor Push Running Pace as workout types. It reframes the core metric rather than bolting a workaround onto the old one, which is a thoughtful piece of design, and per the research above it has gone three hardware generations without a measurable accuracy improvement.
The third piece costs a gym nothing in equipment. The American College of Sports Medicine and NCHPAD run an Inclusive Fitness Specialist Certificate, an online qualification covering adaptive training protocols, Universal Design and accessibility standards, carrying 18 continuing education credits.
Almost none of that is AI, and almost none of it features in the sector's technology conversation, which tends to treat personalisation software as the frontier.
The Human Coach Question
The Les Mills 2026 Global Fitness Report, which surveyed more than 10,000 consumers across five continents, found that only 10% prefer AI-created workouts to human-led ones. As reported by Athletech News, 11% of 16 to 27 year olds and 9% of 28 to 40 year olds favoured AI, against 13% of the over-55s. The most technology-native groups were the most resistant.
A coach who understands your impairment, your classification, your equipment and how your body responds on a bad day is doing work that no current model is trained to do. In many regions that coach may be the only person within driving distance who knows the sport at all. The pathways described in our guide to cerebral palsy sport and to frame running run through people, clubs and classifiers, not apps.
The industry is converging on the idea that technology should take the administrative load so humans can do the relational work. That's a reasonable ambition, and one call out I would like to make is that it has to be built for everyone.
What to Ask Before You Join a Gym
This section is my opinion rather than reporting, so I should declare an interest up front. At the time of writing this post, I work in a technology leadership role for the Y, a not-for-profit community sport and recreation provider, and the Y runs some of the centres I point to below as an example of good practice.
With that on the table, marketing language about inclusion is cheap. These are the questions I would ask, and they are specific enough that the answers give you a better understanding of centre suitability.
Can I get to and onto the equipment? Ask about floor space between machines, whether any strength equipment has a removable seat or a zero starting load, and whether the accessible entrance is the main one.
Who on staff is qualified? Ask directly whether any trainer holds an inclusive or adaptive fitness qualification, such as the ACSM/NCHPAD certificate. "We're happy to help" is not a qualification.
What happens to my numbers? If the gym uses connected equipment or an app that estimates load, calories or recovery, ask whether it has any settings for adaptive users, and treat unadjusted outputs as rough at best.
Is there a program or just access? Access to a room is not the same as a session. Ask whether there is a coached adaptive session, and when.
Can I try before I commit? Any facility confident in its accessibility will let you come and test it.
What a good answer looks like
The Boroondara Disability Sport and Recreation Information Hub in Melbourne gathers what three public aquatic and leisure centres offer into one place. The physical detail is specific rather than aspirational, pool ramps, hoists and entry points with handrails, wide unobstructed pathways inside and out, lift buttons labelled in Braille, and accessible change and shower facilities. Staff can read documents aloud and arrange interpreters. The Boroondara Leisure and Aquatic Facilities are operated by the Y in partnership with Boroondara City Council.
The programming matters as much as the centre layout. The hub lists Come and Try days, WheelTalk sessions and a six-week Blind and Low Vision Swim Program for adults, which is the difference between a centre offering access to a room and a centre offering a session with a coach in it. It runs with support from Together More Active, a Victorian Government initiative, alongside Disability Sport & Recreation, the state body for Victorians with disability, and Blind Sports and Recreation Victoria.
Boroondara publishes its accessibility detail openly instead of making you ring and ask, and it names its partner organisations, which is how I judge whether the expertise behind a program is real. I think both are fair to expect anywhere else.
Governing Bodies and Where to Get Involved
Adaptive sport is organised through national bodies and community programs rather than through fitness technology companies, and these are the organisations to contact.
Australia: Disability Sports Australia, National sports organisation, more than 65 years old and a founding member of Paralympics Australia; runs the Abilities Unleashed program connecting people to local sport.
United States: Move United, National network of adaptive sport clubs and events, searchable by location and sport.
United States: NCHPAD, The National Center on Health, Physical Activity and Disability, based at the University of Alabama at Birmingham; evidence-based programs and resources.
United Kingdom: Activity Alliance, National charity for disabled people's inclusion in sport and activity; publishes the Annual Disability and Activity Survey.
United Kingdom: WheelPower, National charity for wheelchair sport, based at Stoke Mandeville Stadium; over 40 sports, plus free online exercise classes.
Beneath the national bodies sits a state layer, and in Victoria that is Disability Sport & Recreation, which has worked with individuals, clubs, state sporting associations and government for 60 years. It is a registered charity with the Australian Charities and Not-for-Profits Commission and is recognised as a State Sport and Recreation Body. Its WheelTalk disability awareness program reaches more than 10,000 school children a year, and DSR is one of the partners behind the Boroondara hub described earlier.
Internationally, the International Paralympic Committee is the global governing body of the Paralympic Movement and works with more than 200 member organisations. Its National Paralympic Committee directory is the fastest route to the right body in any country not listed above.
Final Thoughts
I don't think anyone set out to exclude anyone here. People built for the users they pictured, and disabled athletes weren't among them. That is a far easier problem to fix than deliberate exclusion would be.
What gets me is that none of this needs inventing. The accessible strength machines are on sale now. The push tracking is already in a watch millions of people wear. The qualification costs a few hundred dollars and a weekend of someone's time. What's missing is the decision to bother, and that is the one thing readers, members and program managers can actually shift by asking for it.
If you're a disabled athlete reading this, the organisations I've listed will get you into a sport faster than any app will, and I'd treat the numbers on your wrist with scepticism until the research catches up. If you work in a gym or a community program, the questions above are the ones your members are already asking quietly.
The technology will get there. It tends to arrive faster when someone names the gap out loud.
Frequently Asked Questions (FAQ)
Q: What is adaptive fitness technology?
A: It is any equipment, software or wearable designed or adjusted so that people with disability can train effectively. In practice it spans three categories: accessible hardware such as strength machines a wheelchair user can roll into, adapted tracking such as counting pushes rather than steps, and the human expertise layer, meaning trainers with an inclusive fitness qualification.
Q: Are fitness trackers accurate for wheelchair users?
A: Not reliably, at least for energy expenditure. A 2024 study in JMIR Formative Research found the Apple Watch Series 4 underestimated energy expenditure by 27.4% during wheelchair propulsion, while the Fitbit Versa overestimated it by 73.9%. Heart rate was better on the Apple Watch, at 8.5% error, though both watches read low. Worth knowing: the study found much the same errors in people without disability doing the same activity, so this is the device struggling with wheelchair propulsion rather than with you. Use the trend over weeks rather than trusting any single day's calorie figure.
Q: Can I use an AI personal trainer app if I have a disability?
A: You can, and some people find the structure genuinely useful, but treat the output as a starting suggestion rather than a prescription. These apps build their recommendations on the same estimated inputs your tracker produces, and those estimates have been shown to be unreliable during wheelchair propulsion. An error in the calorie or recovery figure carries straight through into what the app tells you to do next. Pair it with a coach or physiotherapist who knows your impairment where you can.
Q: Do I need to join a national body to take part in adaptive sport?
A: No, not to start. Community programs and local clubs will welcome you without any membership. National bodies matter once you want to compete, because that is where classification, sanctioned events and pathways sit.
Q: What qualification should an adaptive fitness trainer have?
A: The ACSM/NCHPAD Inclusive Fitness Specialist Certificate is the most widely recognised specialist credential internationally, covering adaptive training protocols, Universal Design and accessibility standards. Many excellent adaptive coaches hold sport-specific coaching qualifications from national bodies instead, which is equally valid.
Q: Is adaptive gym equipment expensive for facilities to install?
A: It varies widely, and this article does not cover pricing. Accessible design often serves other members too. Zero-starting-load machines and removable seats are genuinely useful for older adults, people returning from injury and complete beginners, which is why Universal Design is a core part of inclusive fitness training.
Disclosure: the author works in technology leadership for The Y, a not-for-profit community sport and recreation provider that operates facilities referenced in this article and purchases equipment and software from vendors operating in this market. No organisation mentioned here has reviewed, funded or influenced this piece.
Category: Technology & Sport
Related Posts:
