Making life better for people with disabilities is a laudable goal, but accessibility technology hasn’t traditionally been popular among VCs. In 2022, disability tech companies have attracted around them 4 billion dollars in early-stage investments, which was a fraction of it fintechrecruitment, for example.
One reason is that disability tech startups are often considered too niche to achieve business viability—at least at the scale that venture capital requires. By default, it’s supposed to be building for a minority. However, some startups in the space are also starting to serve the wider population — and throw in some artificial intelligence always helps.
Both cases are a balancing act: The larger business case must make sense without losing sight of the startup’s mission statement. AI, meanwhile, must be leveraged in a non-artificial way to pass the due diligence sniff test.
Some accessibility-focused startups understand these necessities, and their strategies are worth a look. Here are four European startups doing just that.
Visualfy
Image Credits: Visualfy
Visualfy harnesses artificial intelligence to improve the lives of people with hearing loss. The Spanish startup focuses on security and autonomy — it includes a sound recognition AI that recognizes fire alarms and the sound of a baby crying at home. “AI is critical to our business,” CEO Manel Alcaide told TechCrunch last month.
The company offers consumers an app that also serves as a companion to Visualfy Home, the hardware suite consisting of three detectors and a main device. It also entered the public domain with Visualfy Places — no fluke to boot recently raised funding by Spain’s national state railway company, Renfe.
One reason Visualfy is gaining traction on the B2B side is that public spaces are required to provide accessibility, especially when health and safety are on the line.
In an interview, Alcaide explained that the devices and PA systems that Visualfy will install in places like stadiums could also monitor air quality and other metrics. In the EU, meeting these other targets could help companies get subsidies while doing the right thing for deaf people.
The latter is still very much on the mind for Visualfy, which is incorporated as a B Corp and employs both hearing and non-hearing people. Inclusion of deaf people in all steps is an ethical stance – “nothing for us without us”. But it’s also common sense for better planning, Alcaide said.
Knisper
Image Credits: Audus Technologies
People with total hearing impairment are a smaller part of a large and growing group. By 2050, 2.5 billion people are predicted to have some degree of hearing loss. Due to a combination of reasons such as stigma and cost, many will not wear hearing aids. This is the audience Dutch B2B startup Audus Technologies is targeting with its product, Knisper.
Knisper uses artificial intelligence to make speech easier to understand in environments such as cinemas, museums, public transport and work calls. In practice, this means separating the audio and mixing it into a cleaner track. It does this without increasing ambient noise (something not every hearing aid company can say), making it comfortable for anyone to listen to, even without hearing loss.
A former ENT doctor, Audus founder Marciano Ferrier explained that this could not be achieved with similar results before AI. Knisper was trained on thousands of videos in multiple languages, with variations such as background noise and distorted speech. That took work, but Audus is now leaving the development stage and focusing on adoption, CEO Joost Taverne told TechCrunch in February.
“We already work with several museums, including the Museum of Fine Arts in Boston,” said Taverne, a former congressman and diplomat who spent time in the US. audiobook of the diary of anne frank accessible for people with hearing loss. And now we have the solution for the workplace.”
Going to the B2B market is not an easy path, so it makes sense for Audus to focus on customers such as museums. They are often noisy, which can make audio guides difficult to hear. Using Knisper’s technology to make them more understandable brings benefits to the general public, not just those with hearing loss, which facilitates adoption.
Whispp
Image Credits: Whispp
Fellow Dutch startup Whispp it also focuses on speech, but from a different angle. As TechCrunch reported from CES earlier this year, its technology is transforming whispered speech in a natural voice in real time.
Whispp’s primary target audience is “an underserved group of 300 million people worldwide with vocal impairments who have lost their voice but still have good articulation,” web page explains.
For example, people with voice disorders that leave them only able to whisper or use their esophageal voice. or who stutter, like CEO Joris Castermans. He knows very well how his speech is less affected when he whispers.
For those with impaired articulation due to ALS, MS, Parkinson’s, or strokes, solutions such as text-to-speech apps already exist — but these have drawbacks, such as high latency. For people who are still able to articulate, this may be too much of a compromise.
Thanks to audio-to-audio artificial intelligence, Whispp is able to provide them with a voice that can be produced in real time, is language agnostic and sounds real and natural. If users are able to provide a sample, it can even sound like their own voice.
Since there’s no text in the middle, Whispp is also more secure than alternatives, Castermans told TechCrunch. That could open up use cases for non-silent patients who need to have confidential conversations, he said.
How much non-voice-impaired people would be willing to pay for Whispp’s technology is unclear, but it also has several monetization avenues to explore with its core audience, such as the subscription it charges for its voice calling app.
Acapella
Image Credits: Acapela Group
Whispp highlights the need some have to store their voice for future use. Known as voice bankingthis process is what Acapela hopes to facilitate with one of its services launched last year.
Acapela Group, which was bought by Swedish accessibility technology company Tobii Dynavox for €9.8 million in 2022located in the text-to-speech space for several decadesbut only recently artificial intelligence has changed the picture for voice cloning.
The results are much better and the process is also faster. This will lower the bar for voice banking, and although not everyone will do it yet, there may be demand for people who know they are at risk of losing their voice after being diagnosed with certain conditions.
Acapela does not charge for the initial phase of the service, which consists of recording 50 proposals. Only when and if they need to install the voices on their devices do users need to purchase it, either directly through Acapela or through a third party (partner, reseller, national health insurance plan or other).
In addition to the new potential that AI unlocks, the above examples show some routes that startups are exploring to expand beyond a core target of users with disabilities.
Part of the thinking is that a larger addressable market can increase their future revenue and spread costs. But for their customers and partners, it’s also a way to stay loyal to definition of accessibility as “the quality of being accessible or usable by everyone, including people with disabilities.”
