An Abu Dhabi start-up called Limb is developing an AI-powered app to give people feedback while they do physiotherapy exercises at home. Its approach also tackles a specific blind spot in computer-vision training data: movement tracking can struggle with the loose-fitting clothes many patients in the UAE wear.
Watch Desk analysis
What happened
Limb’s app sends patients exercise programmes created by physiotherapists, using their own demonstration videos. As patients exercise, computer vision tracks key points on the body and compares movements with the prescribed exercise; the app can then give spoken feedback. Physiotherapists can see whether exercises were completed and review reported pain scores.
Co-founder Shamma Alsaedi told Gulf News that a survey of 80 UAE physiotherapy patients, conducted as part of her master’s research, found 71 per cent did not do their home exercises consistently or at all. That is a small survey, not a measure of every patient’s habits, but it helps explain the problem Limb is trying to address.
The company estimates that around 90 per cent of publicly available exercise datasets show people in fitted sportswear. Limb says it is collecting movement data from people wearing loose clothing, and that its app works in Arabic and English. The company says video is processed to extract movement data, then discarded, and patient data is stored in the UAE. Limb is bringing clinics into an early-access programme; it is not presenting the system as a replacement for physiotherapists.
Why it matters
Home exercises are a routine part of recovery, but patients do them away from the clinician who prescribed them. Limb’s proposed feedback loop could help practitioners see what happens between appointments, while its local data collection addresses a practical question for computer vision: does the system track the people who are actually meant to use it?
There is a meaningful gap between an early-access product and proven improvements in recovery. Better tracking is not, by itself, evidence that patients heal faster or that feedback is always right. The useful next evidence will be how the system performs with real patients and whether clinicians find it helpful in practice.
Our read
The strongest idea here is not that an app can replace a physiotherapist. It is that a model trained around everyday users might help a physiotherapist stay involved when the patient is back at home. That is a more grounded ambition than giving a camera a medical-sounding job title and hoping for the best.
Limb’s focus on clothing, language and elderly users is worth watching. The next step is evidence from the early-access programme: accuracy, clinician oversight and whether patients actually keep doing the exercises.
What to watch
- What results Limb shares from its early-access programme with clinics.
- How reliably its computer-vision system tracks movement in loose-fitting clothing.
- Whether clinicians find the feedback useful and maintain oversight of patient programmes.
Discussion spark: Would you trust AI-generated movement feedback to support home physiotherapy if a clinician remained responsible for the programme, or should that feedback stay strictly advisory?
Sources and evidence
- Abu Dhabi Start-Up Transforms Physiotherapy with AI-Driven Home Recovery Solution (12 October 2026, 09:38 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.