The UK Assistive Technology Market in 2025: Where the Gaps Are

The UK assistive technology market is not underfunded. It is misallocated.

Money flows to hardware with clinical procurement pathways, and away from software that would reach more people for a fraction of the cost. This post maps where the demand actually sits, why British Sign Language keeps losing to American Sign Language in product roadmaps, and why the June 2025 regulatory deadline changes the buyer, not just the rules.

The addressable population

Start with who we are talking about.

Group	UK population	Source
Deaf or significant hearing loss	~2 million	RNID
BSL as a preferred language	87,000–151,000	British Deaf Association / 2021 Census
Registered blind or partially sighted	~340,000	RNIB
Living with sight loss (broader)	~2 million	RNIB
Both hearing and sight loss (deafblind)	~450,000	Sense

Two numbers in that table matter more than the rest.

87,000 to 151,000 BSL users is the number that gets a product killed in a planning meeting. It reads as small. It is used to justify pointing a roadmap at ASL instead, where the equivalent figure is roughly half a million in the US alone.

450,000 deafblind people is the number almost nobody builds for, and it is the one that should govern design. A tool that assumes the user can see the screen or hear the output has excluded them entirely. Design for that constraint and the resulting product is more usable for everyone else — which is the argument we make at length in a separate post.

Market size, and what the figure hides

The global assistive technology market sits in the region of 26 billion US dollars in 2025, with projections in the 7–8% CAGR range through the early 2030s. The UK slice trends toward roughly £4.2 billion by 2030.

That headline number is mostly not software. Break it down and the pattern is clear:

• Mobility and daily-living aids — wheelchairs, hoists, adapted equipment. The largest slice by revenue.
• Hearing devices — hearing aids and cochlear implants. Well-established clinical procurement, mature manufacturers.
• Vision aids — magnifiers, screen readers, and a small premium hardware segment.
• Communication software — the smallest slice, and the fastest growing.

The revenue concentration follows procurement, not need. Devices with a clinical pathway get bought by institutions. Software without one gets bought by individuals, who are, on average, poorer than the general population: disabled households in the UK face substantially higher living costs and lower median income.

That mismatch is the entire commercial thesis for a mobile-first tool. The people with the greatest need sit in the segment with the least purchasing power, which means anything priced like clinical hardware structurally cannot reach them.

What the current products cost

The two leading vision products in the UK market:

OrCam MyEye — a camera unit that clips to a spectacle frame, reads text, and recognises faces and products. Around £4,000–£6,000 depending on configuration. Excellent at what it does. It is also a second device to buy, charge, carry, and eventually replace.

Envision Glasses — Google Glass hardware running Envision's software. Around £2,500 and up. Strong text recognition, dependent on a discontinued hardware platform.

For sign language, the picture is thinner still. Most consumer sign-language apps are ASL fingerspelling teaching tools. They teach the alphabet. They do not perform real-time recognition of lexical signs, and they are not BSL.

The relevant comparison for a deaf patient in an NHS waiting room is not another app. It is a booked human interpreter, at roughly £45–£90 per hour plus travel and a minimum call-out, typically requiring days of notice. Interpreters are better than any software at genuine conversation, and they should be funded. But they cannot be summoned for the unscheduled ninety seconds when a receptionist asks which GP you are here to see.

That ninety seconds is the gap.

Why BSL keeps losing to ASL

This is a structural problem, not an oversight.

Data availability. The large public sign language datasets are overwhelmingly ASL. WLASL, MS-ASL, and How2Sign are all American. A team that wants to ship sign recognition next quarter uses what exists. BSL datasets are smaller, more fragmented, and frequently licence-restricted.

Market sizing done lazily. A US-first ASL product addresses a larger single-language market. BSL looks like a rounding error next to it — until you notice that BSL, Auslan, and New Zealand Sign Language form the BANZSL family and share substantial lexical overlap, which makes the real addressable population considerably larger than the UK figure alone.

A genuine linguistic difference. This is the part most often misunderstood by people outside the field: BSL and ASL are not dialects of each other. They are unrelated languages with different grammars and different manual alphabets — BSL fingerspelling is two-handed, ASL one-handed. ASL descends from French Sign Language. A model trained on ASL is not "mostly right" on BSL. It is wrong.

The result is that BSL users get products designed for a different language, and the products underperform, and the underperformance is read as evidence that the market is not there.

The regulatory shift

The European Accessibility Act obligations took effect on 28 June 2025. It requires that a broad class of products and services — e-commerce, banking services, e-books, transport ticketing, computers, smartphones, self-service terminals — be accessible to people with disabilities.

The UK is outside the EU, and the EAA does not apply domestically. That fact matters less than it appears, for three reasons:

1. Extraterritorial reach. Any UK business selling into the EU market falls in scope. That is a very large number of British companies.
2. Standards gravity. EN 301 549, the harmonised European accessibility standard, incorporates WCAG 2.1 Level AA. UK public sector bodies are already bound to the same WCAG baseline by the 2018 Public Sector Bodies Accessibility Regulations. The technical target is converging regardless of jurisdiction.
3. The buyer changes. Compliance moves accessibility from a discretionary CSR line into a procurement requirement with a named owner and a budget. That is a different sales conversation entirely.

Alongside it, the British Sign Language Act 2022 gave BSL legal recognition as a language in England, Wales and Scotland, and obliges UK government departments to report on their use of BSL in public communications. It carries no direct enforcement mechanism, but it establishes an expectation — and expectations are what procurement teams document against.

Where the actual gap is

Putting it together, the underserved position looks like this:

Mobile-first, not hardware-first. The device is already in the user's pocket. Every product that requires a purchase before first use loses most of the people who most need it.

BSL-native, not ASL-ported. Trained on British Sign Language from the ground up, with a roadmap toward BANZSL coverage rather than a US pivot.

On-device, not cloud. Three reasons, in order of weight: privacy — a camera pointed at a medical conversation must not stream anywhere; latency — conversation does not tolerate a round trip; and availability — hospital corridors, the Underground, and rural classrooms have no usable connectivity, and those are exactly the places the tool is needed.

Institutional distribution, individual use. The individual cannot pay £4,000. The NHS trust, the local authority, and the multi-academy trust can pay for a deployment — particularly when the alternative line item is interpreter booking costs, and particularly when a compliance obligation has a name attached to it.

What we do not know yet

We are a small team making a market bet, so it is worth naming what remains unvalidated:

• Whether institutional buyers will procure software without a clinical device classification is genuinely unproven at our scale.
• Whether 93 signs — or 500 — reaches the threshold of usefulness that makes a deaf user open the app a second time.
• Whether on-device accuracy holds up across the demographic breadth of real deployment rather than our contributor pool.

Our Q4 2025 pilot with three UK special needs schools is designed to answer the second and third of those. The first is a fundraising question, and it is the reason we are talking to Innovate UK.

---

Figures cited are from RNID, RNIB, Sense, the British Deaf Association, and the 2021 Census. Market projections are directional estimates, not audited forecasts — if you have better data, we would genuinely like to see it: hello@devsandvisuals.com.