... misleading title to the article there. While it's £40 (or is it twice as much for 2?) for the cameras, it's going to take a lot of time gluing them to a pair of glasses. :D
Neurotechnologists in London have invented a £40 device that lets you control computers with your eyes. Costing hundreds of times less than existing 3D eye tech it could revolutionise the life of the disabled say the team at Imperial University, allowing users to direct a wheelchair by simply looking at where they want it to …
Not really misleading, they are showing the cost 'off the shelf', if you were to buy direct and have it constructed offshore, the cost would probably be similar, at the very least a finished device would be sub £100, which would make a massive difference to disabled people, and of course anything that improves independence for the disabled lowers the cost for the able bodied masses that pay for everything.
About bloody time. my kid has CP and a basic eye tracking system, that doesn't even work as a general purpose mouse type device, costs in the region of £10,000. I can employ a whole human for 4 hours a day for a year at that rate! (and in fact, his school do just that)
I had a crack at doing this with glasses mounted cameras - the tricky bit is to work out where the head is pointing and subtract the difference - otherwise, my 'hacked together in my back shed' prototype actually worked quite well, provided you didn't move your head! You don't /really/ need to track both eyes, though if you have normal vision it'll give you some depth clues. Nystagmus sufferers are more likely to get problems if their vision is partially occluded, so two cameras might set it off worse than one. Anyway, that's why the good lord invented Kalman filters.
Pity the research is not fully open. I know a few people at reading makerspace who are involved in
building tech for the disabled/blind/... who may be interested in replicating this.
With the lowering of video card GPU's its possible to do the sort of calcs via a battery powered device previously thought impossible or not cost effective.
First a thumbs up to the researchers.
But the article is a fail.
Comparing a portion of the cost of a system against the entire system is very misleading.
This camera has to tie into a computer system which isn't part of the 40 pound cost. Then there is the cost of the software... So you still end up with an expensive system.
Sorry but the software isn't free.
I'm sure I will get a lot of down votes by commentards who haven't thought this through.
No, I didn't miss that point.
Your cost in hardware isn't actually that trivial but the bulk of the cost is going to be in the software.
(Which you belatedly admit ;-)
You really don't expect anyone to run a wheelchair using this as input on completely open source with no proprietary software, did you?
Or the liability insurance you'll need in case something happens to someone like driving over a cliff, or someone hacks their computer... (Or the wheelchair suddenly lurches in to traffic and the poor sod gets hit by a bus or garbage truck [lorry]...)
You also have to realize that the use of this would be a very small niche environment. So you have a small user base to cover the cost of software/hardware/insurance/etc ...
Like I said, I applaud the tech, spot on and with more research could lead to some other interesting devices. However... at best you could probably reduce some of the total costs but not by as much as you think...
Now if you had said shoot the lawyers first... then maybe you'd have a chance.
> ...but the bulk of the cost is going to be in the software.
That's very 20th century thinking.
I think that you will find that the replication cost of the software is pretty minimal.
They have done the grunt work already for the research.
The per unit cost is what the are talking about here and that is hard to reduce to zero.
A mate of mine was telling me about when he was learning to paraglide. He kept crashing into trees when trying to land, even if there was only one tree for miles around. Eventually he realised that it was because he kept looking at the tree: the paraglider was 'steered' in the direction he was looking because of subtle body movements associated with turning the head. Once he started making a conscious effort to ignore the tree and look only at a patch of empty field he was able to make perfect landings.
So yes, steering a wheelchair by looking where to go is potentially dangerous as people tend to look at obstacles and hazards.
Biting the hand that feeds IT © 1998–2019