Blind and Visually Impaired Community
↑
6
↓
Designing a smart guiding robot for the visually impaired (self.Blind)
submitted by galani3
We are currently trying to design a smart guiding robot for my senior design class and wanted to get feedback on which functionalities or designs do you think would greatly help the visually impaired navigate through buildings? Any information will be greatly appreciated!
rkingett 3 points 5y ago
yawn. Another navigational aid. I do have to admit though, this robot would be way more interesting. If it is voice activated though I can't use it because I stutter but the post below outlined the 2 main things.
galani3 [OP] 1 points 5y ago
What type of interface would you suggest be the best?
rkingett 1 points 5y ago
If it does have to be just voice activated, Is there a way where I can text a number unique to me, and it would convert it to text to speech so the robot could hear it?
what about an IFTTT trigger that would allow me to, say, send a SMS message to make it do certain things if there will be IFTTT support?
Or can you have it like the latest SIRI? where you type to it and it replies verbally?
Most people would want the voice activated thing, so can you make the interface a toggle at will?
But in reality, we need someone to make self driving skates or a self driving bike. Not another "lets help you get around!" thing.
what about an IFTTT trigger that would allow me to, say, send a SMS message to make it do certain things if there will be IFTTT support?
Or can you have it like the latest SIRI? where you type to it and it replies verbally?
Most people would want the voice activated thing, so can you make the interface a toggle at will?
But in reality, we need someone to make self driving skates or a self driving bike. Not another "lets help you get around!" thing.
KillerLag 3 points 5y ago
Would this be a robot the person would put their hand on, and it would move to the necessary location? Or would it be closer to a system that would tell the person which direction to go (Left, right, 2 o'clock, etc).
Two things it would definitely need to do is check for drop-offs and detect obstacles in the way.
Two things it would definitely need to do is check for drop-offs and detect obstacles in the way.
galani3 [OP] 1 points 5y ago
We were thinking the robot would automatically attach to the bottom of a user's current cane at an angle. It would then be the one guiding you, so you just follow it. Thanks for the feedback!
KillerLag 1 points 5y ago
A few issues with that design. If it attaches to the tip of the cane, then that would make the cane much more difficult to use, if not outright impossible to move. The main purposes of a cane is to check for things coming up, and to provide safety.
If the robot holds the cane stationary, then the things can impact the person from the sides (think of chairs, benches, mail carts, doors, etc).
If the robot moves with the cane tip.... well, let's assume the robot is only one pound. Attach that to a 54" pole (the average height of canes for the population). Now, for the two main techniques (constant contact and two point touch), you should be scanning from side to side for each step (so one step, one scan across, from shoulder to shoulder). Try walking at a normal pace and see if you can move that weight at that speed. Cane tips can do it because a) the larger ball tips are hollow and b) the denser tips are relatively small and lightweight.
Additionally, a design that is focused on the cane would not be helpful to a guide dog. If the robot was about, say... the size of R2D2, that is large enough for the person to give the dog a command to follow (hopefully the dog would follow, I don't believe any dog has ever been trained to follow a robot before) and also for a person to put their hand on the robot and walk with it.
If the robot holds the cane stationary, then the things can impact the person from the sides (think of chairs, benches, mail carts, doors, etc).
If the robot moves with the cane tip.... well, let's assume the robot is only one pound. Attach that to a 54" pole (the average height of canes for the population). Now, for the two main techniques (constant contact and two point touch), you should be scanning from side to side for each step (so one step, one scan across, from shoulder to shoulder). Try walking at a normal pace and see if you can move that weight at that speed. Cane tips can do it because a) the larger ball tips are hollow and b) the denser tips are relatively small and lightweight.
Additionally, a design that is focused on the cane would not be helpful to a guide dog. If the robot was about, say... the size of R2D2, that is large enough for the person to give the dog a command to follow (hopefully the dog would follow, I don't believe any dog has ever been trained to follow a robot before) and also for a person to put their hand on the robot and walk with it.
RogueLeaderJ 1 points 5y ago
My wife has RP and while she isn't blind yet, she has really bad blind spots and acute tunnel vision.
Honestly, I think what would help her the most is a system similar to a self driving car. Something that could tell her if there was something on the floor in front or behind her (like a FLOOR IS WET cone). Also, fast moving objects coming from all directions. Also, with RP, low light conditions are a huge concern.
Honestly, I think what would help her the most is a system similar to a self driving car. Something that could tell her if there was something on the floor in front or behind her (like a FLOOR IS WET cone). Also, fast moving objects coming from all directions. Also, with RP, low light conditions are a huge concern.
EndlessReverberation 1 points 5y ago
Yeh, to be honest, I'm sorry but this robot is a terrible idea. I would not be as blunt if you were simply asking if the project was worth pursuing, but more often then not sighted people who do these kinds of projects just assume they know what blind people need; such sighted people are almost always wrong. Next time reach out to the blind community before you have settled on an idea. If you got this idea from a blind person, I'm sorry, it's not your fault, they just gave you a bad idea.
Would it be possible to make a small device that a blind person could pick up, choose where they were headed, and then receive real-time audio or tactile feedback for which way to go, which turns to make etc. In other words, the main issue with your idea is the guiding/dragging blind people around part. A system for indoor maps/directions, would be interesting.
Would it be possible to make a small device that a blind person could pick up, choose where they were headed, and then receive real-time audio or tactile feedback for which way to go, which turns to make etc. In other words, the main issue with your idea is the guiding/dragging blind people around part. A system for indoor maps/directions, would be interesting.
galani3 [OP] 1 points 5y ago
The robot is not a set and stone design. That is why we're trying to get as much feedback before settling down on a certain design. Thanks for the feedback
-shacklebolt- 1 points 5y ago
I'm just thankful this is some theoretical "design class" project.
What a terrible idea, especially putting it at the bottom of the cane (the hitting things part!)
What a terrible idea, especially putting it at the bottom of the cane (the hitting things part!)
galani3 [OP] 1 points 5y ago
As mentioned this is a "design class". The robot design was one idea that was presented to us. We want to get as much feedback to help design a better device.
This nonprofit website is run by volunteers.
Please contribute if you can. Thank you!
Our mission is to provide everyone with access to large-
scale community websites for the good of humanity.
Without ads, without tracking, without greed.
scale community websites for the good of humanity.
Without ads, without tracking, without greed.
©2023 HumbleCat Inc • HumbleCat is a 501(c)3 nonprofit based in Michigan, USA.