The white cane (and its many variants) is an everyday carry for many visually impaired people. This low-tech tool allows those afflicted by visual impairment to safely navigate the world around them, and has been ubiquitous in many parts of the world for decades. [Madaeon] has been hard at work going one step further in prototyping an open-source assistive wearable that could help in situations where a cane is not practical, or useful.
The T.O.F Wristband V2 alerts its wearer to nearby obstacles through vibrations, and is able to detect objects up to four meters away. As the wearer veers closer and closer to an obstacle, the vibration increases in frequency. A time-of-flight distance sensor is controlled by a Feather, and the whole system is powered by a small lithium-polymer battery. The prototype consists of just four components plus a 3D printed case and bracelet, which inevitably keeps down costs and complexity.
Version two of this project picks up where version one left off. In that project, [Madaeon] mentioned the possibility of squeezing this project down to the size of a ring. Perhaps with better battery technology, a ring-sized sensor might just be possible one day.
This isn’t the first wearable that has set out to assist the visually impaired. Back in 2019 we covered a laser-augmented glove that attempts something very similar.
By some estimates, nearly one billion people worldwide have some degree of visual impairment. Assistive devices like the T.O.F Wristband V2, and others like it, offer these people the potential for greater independence and an improved standard of living.
[youtube https://www.youtube.com/watch?v=p_S53d9pIxo%5D
This is not a substitute for a cane. Aside from the possible detection of objects at head height this lacks many of the features of a cane. A cane identifies the user as visually impaired by its distinctive appearance. A cane is as much a sonic device as tactile. The tapping sound helps the user identify the surface, grass, asphalt, gravel, concrete. Skilled users interpret the echo as well. You can interpret the texture of the surface. The cost is minimal. There are no batteries to charge. A cane is a very difficult item to improve upon. Many have tried. In addition to a cane this could possibly be of use.
This!
I am visually impaired. I have a cane (I don’t use it yet), I know how to use it. And I laugh at every attempt to “improve” it. How this device would help someone blind avoid falling down the stairs or keep him/her from tripping on something, like a single step, or any of temporary obstacles? I was once tripped by extension cord stretched across a sidewalk by some idiot. White cane works in any of these scenarios flawlessly…
I’d love to see a project that uses augmented reality to help someone like me – a person who can see but sometimes has problem seeing things in certain light conditions. For example I know all signposts on the sidewalks I frequently use because I banged into them on various occasions. In certain conditions silver painted post just disappears in the background of gray concrete sidewalk. For these situations I envision a mode that would add contrasting wireframes to all edges and obstacles. And when it gets dark, a night-vision function would be great for me…
Just a thought, for your problem distinguishing light posts. What if you used a bright/coloured flashlight/flash unit to light up the posts? It would allow for greater contrast between the posts and the sidewalk, it would look weird I suppose but if it helped you navigate your environment better who cares?
The problem occurs usually when it’s overcast in the middle of the day. I think the flashlight would have to be rather strong to work in these conditions. Fortunately this kind of situation happens rarely, as I have developed over the years a very good spacial memory. Besides someone who accidentally ends up hugging a lamp post or a street sign to other people looks like a drunk. Someone using a strong flashlight in the middle of the day will appear to others as someone who is nuts…
Well you could tell people you’re Diogenes. :)
I give him credit for the effort. This area needs a lot more effort and investment. But, simplicity, failsafe reliability, cost, learning curve, ergonomics and other factors make it a far more difficult problem than at first glance. The cane works. 15 bucks US. Next option is a dog. Only suitable for some people. Expensive to keep, select, train and train the user. Thousands of dollars. Many thousand. In that gray area between a cane and a dog, many have tried and few have come up with anything actually useful. There are not enough impaired people to make a profitable market and attract the necessary R&D dollars. As ever, it’s about money.
There is plenty of money in this, because in most EU countries state refunds at least a part of the cost of any device or material and treatment used by someone impaired. Some countries pay more, some pay less. And some, like mine, have some crazy rules about it. For example if someone had his leg amputated, here he would get refunded (IIRC) 95% of the cost of prosthetic replacement (but not the expensive one) and the same goes for fitting it to the stump. However first fitting is to form the stump, and is good for only few months. But one can’t get another fitting refunded before 3 years from the last one. So one gets a prosthetic leg and can use it only for few months, and then he would have to pay for new fitting or wait 2,5 years or longer.
Few years ago I wanted to get a microscope or other magnifier to help me soldering smaller parts. I requested a refund (it’s done before the purchase) from National Health Fund (state-owned health insurance provider). They refused on the grounds that I’m too blind to solder, so I don’t need it…
“For these situations I envision a mode that would add contrasting wireframes to all edges and obstacles. And when it gets dark, a night-vision function would be great for me…”
Given the advances in sensing technology and computing, I’d think that such a thing is do-able. I wonder if a cellphone could be leveraged to do, or help, with the latter ? Getting wireframes, bounding boxes, etc is already a known art on even a RPi. Getting the pertinent info to you is, perhaps, more problematical. But in your specific case, since you retain some vision, perhaps something like Google Glass could be used. Or something from another of the other AR “smart glasses” manufacturers. Magic Leap comes to mind as something that purported to already do most of what you’ve asked for.
I don’t know what happed to it, or came from it, but there was a HaD project a while ago that, IIRC, was along this line. Let me see if I can find a link.
https://hackaday.io/project/27111-digital-white-cane
OK, they were haptic feedback. Interestingly under the same (generic) name, Digital White Cane.
Google Glass actually gave me this idea. The technology exists, but no one implemented it yet. There is a device, that looks like one of those VR sets from 1990s, that has a camera on the front and could work as magnifier and color remapper (for readability), but it couldn’t be used outside. And it was big and heavy…
How about adding sensors to end of cane? Sensing surfaces reflectivity, color, perhaps temperature, … Or low tech amendment – adding small magnet to cane tip would allow wielder to detect if the tip is near iron/steel object.
This is not necessary, as a skilled user can feel the texture of the ground. Besides, there are different tips for the cane. There is for example a roller tip for those who prefer swiping instead of tapping, and special bent one for the off-road, to feel for brushes and twigs. One thing that is really surprising about the cane, is how versatile it is, considering it’s jut a segmented pipe with a tip and a handle…
Not to mention that the cane also detects holes, and other kind of hazards beyond “obstacles”
I can see where you are coming from (and other commenters have said much the same). This won’t replace a white cane, but projects like this have the potential to tease out new solutions that could go even further to help the visually impaired.
I have often thought about such a device that could help a former blind friend.
The idea was to cause a laser beam to scan the ground.
This way we get a straight line without obstacles.
If an obstacle appears, a niche is formed on the straight line scanned by the laser.
A device detects this crenelation and transforms it into a sound in a pair of headphones.
This sound is synthesized and rendered stereo and by a phase shift effect it locates the location of the obstacle by a sound that locates the direction of the object (stereo).
The volume of the sound gives an idea of the distance while the height gives an idea of the width of the obstacle.
The person directs the module like a walking stick and therefore has an idea of what is in front of him.
The extreme sensitivity and analysis of blind people might allow them to capture many details of their environment.
Translated with http://www.DeepL.com/Translator (free version)
I was thinking the same thing. This is a great project and an interesting start, but to really do this right you’d need the situational awareness of a Tesla coupled with a method of getting all that information back to the wearer in a meaningful way, in a package that weighed nothing, ran for at least a full day and was mostly unobtrusive.
Tall order indeed.
This is interesting but I agree with the previous posts. Maybe this another step towards something wearable……like Geordi’s visor.
my mind immediately went to the same criticisms as other commenters, but i think a lot depends on the implementation. on how processed the signal is. i’m struggling to find the right analogy…
i’ve got an audio program i made for tuning pianos, it has a bunch of limitations but it’s designed as an augmentation to the ear. it filters the signal a little bit (band-pass filter) and then basically raster scans the signal with the horizontal retrace tied to the target frequency…so it’s messy, and it has a bunch of useless information. background noises and harmonics show up, and it shows you the phase even when you don’t care about it. and it doesn’t tell you what every other tuner program does…my app doesn’t tell you what note it hears, it doesn’t tell you sharp or flat, or by how much. it just renders the sound in a way where you get a bunch of extraneous information, from which an astute observer can discern what they need.
so mine really augments your ear, and leaves all the interpretation to you. but the traditional ones attempt to solve the whole puzzle within the computer, and just tell you the interpretted useful data.
i really don’t know how this device works but i can imagine a similar device that doesn’t hardly process the data, that exposes you to a noisy and immediate rendering of its sensor values, and leaves interpretation up to the user. maybe even an entirely analog system, akin to the classic metal detector. i think that would have a lot of potential, in that users would learn intuitively and could discern complicated information that isn’t immediately obvious in the output. if it tries to process the sensor data until it knows the distance and type of the next obstacle, i think it’d be useless. but if it gives you a sound that is idiosyncratically related to the distance, texture, and alignment…i have a lot of faith in a user’s ability to tease out the details, at least once you work out the kinks.
Whats the range? Ive been troubled to find an affordable sensor past the 15/20 meter range..
Forrest Mims created an aid for the blind in the sixties, infrared LED(s) sending out a signal, and getting a return, they fitted eyeglass frames. He created it because a relative was blind, but I don’t remember how well it worked or was accepted. The LEDs were very expensive at that point (and I suspect they were infrared laser diodes, he wrote about those a bit later).
But yes, like another recent post, a good idea isn’t the same thing as knowing the territory. I’m certainly aware that the cane is obviously for the blind, a good marker.
I’m just wondering why not mount this sort of vibration distance sensor in a white cane. Although I’m fairly sure that’s been done before.
If this sort of thing was coupled with aa low powered t.e.n.s. machine and then made into a full body suit, say a sensor every half a foot square, I wonder if the mind would happily rewire itself to accommodate being able to see at a distance with the entire surface of the skin.
The human mind/brain would happily rewire itself to use the surface of the body as substitute for sight, hearing or possibly even taste and smell. It would only take a couple of weeks for the rewiring to happen and has been proven by science. This dude had a curaiss/chestpiece with little cellphone vibrators at various places on it connected to an app on his phone. The app would trigger vibrator combos and provide him with multiple choices of words with 1 of them being correct. It only took about 2 weeks for his brain to rewire to accommodate the vibrations from the vibrators as a substitute for hearing. Saw it on the discovery channel i think. Im pretty sure it was a documentary about neural interfaces.
Now this project or a sonar type setup coupled chestpiece or fullbody suit like what that guy used in the documentary could provide a pseudo-sight type deal that can hilight corners and hard objects where the brain processes the input in a way the user sees a line drawing like shape of the object with shading variations depending on the hardness. It would be very rudimentary in terms of sight and unable to see through glass but it would allow “Seeing” things such as walls, tables, steps, doors and more.
A couple years ago I experimented with the same sensor for the same purpose. It worked fine inside but once I took it outside it became just about useless. Apparently the ambient light saturated it. I tried using sunglass lenses and hollow tubes to restrict the ambient light but never found a suitable fix. The manufacturer’s data sheet and app notes also show decreased range with ambient light. It’s unfortunate because this sensor worked fine inside and was cost effective.