Can the crowd contribute to accessibility?
The crowd appears to be particularly helpful where simple but time-consuming tasks are involved. You can see this in Wheelmap, for example.
Wheelmap
On Wheelmap, persons can quickly and easily rate how wheelchair accessible a particular location is. persons could have been sent out with a page-long checklist, decent training in DIN 18040, and a tape measure. But then the crowd would have turned into a heap. For a city like Bonn, it would probably take ten years to assess a critical number of buildings. Nothing against valid data, it would be fantastic if we knew exactly how accessible all publicly accessible facilities are. The only problem is that in the time it would take to do this, a lot of the buildings would be remodeled, cafes would close, stores would close and new ones would open. It would also be interesting to know how accessible public buildings are. Only: how often do persons go to the citizens' office or the town hall?
That's why bad data is better than none at all. Anyone who has used the application to find a supposedly accessible place and finds out on the spot that the rating is wrong can correct the information, which is also crowdsourcing.
Subtitles
A service called Amara allows persons to crowdsource subtitles for videos on YouTube. A similar service is Opensubtitles.org.
The service YouDescribe allows the creation of audio captions for YouTube clips.
Images for the blind
Images for the Blind creates image descriptions for blind and visually impaired persons.
Improvements for websites
However, Crowd can also be used to improve the accessibility of websites. For example, the discontinued Firefox extension WebVisum allows you to add information such as image or form descriptions to websites so that they can be used more easily by blind persons. Many will object here that such descriptions cannot have the quality of professional services. Against it can be argued that a mediocre description is still better than none at all. After all, there is the possibility here that persons with and without disabilities exchange ideas in order to jointly improve the quality of the descriptions. The non-disabled are thereby incidentally sensitized to possible barriers and the effects of disabilities. Unfortunately, the service no longer exists, but technically something like this could be implemented well.
Be my Eyes
The app BeMyEyes allows blind persons to organize volunteer sighted help via their smartphone camera, for example with shopping or orientation.
It is virtually impossible to get a large group of persons with disabilities to test a website for accessibility. The problem is that there is a wide range of disabilities, and within these groups, the impact of the disability and technical abilities vary widely. In the case of visual impairments alone, there are countless variations. With normal resources, it is practically impossible to map this range in usability or accessibility tests.
So why not enlist the crowd to test for accessibility? This has two advantages: Tendentially, persons are more likely to be reached who will use the service anyway and are therefore interested in its accessibility. In addition, persons sit in their familiar surroundings, use the technologies they use anyway, and thus approach the task in a much more relaxed manner than they would in a usability lab. With a small test group, it is often difficult to tell whether there is actually a problem at a particular point or whether it is just the person in question who cannot cope with it. If, on the other hand, a larger number of persons independently find a problem at a certain point, it is also a problem. Of course, the aim is not to make a completely accessible offering accessible; the crowd can't help with that either. Rather, the crowd is supposed to check a website that has been designed according to the rules of WCAG 2.0. The crowd should detect possible problems, not compensate for the inability of web developers.
Summary
Now that was just a brief overview of services that I am aware of. The potential is huge, the only thing missing is someone who can tap this potential. For example, some screen readers mispronounce words. Here it would be an advantage if users could correct the pronunciation and share these corrections directly via a server. To prevent stupid corrections - some persons just get bored - statistical methods can be used so that a change only becomes valid when it has been implemented by enough persons.
The beauty of the crowd is intrinsic motivation. It's about persons being happy to help other persons without expecting anything directly in return. The offerings are relatively low-threshold, no one has to read up on WCAG or other complex regulations, but he or she can jump right in and get started.
The exciting thing about crowdsourcing, in my opinion, is the proactive approach. The disability scene in Germany is too passive for my taste. Crowdsourcing gives them the tools to improve things themselves - just like Wheelmap.