for some reason, glovepie came to mind.
dont know why though…

Posted at 4:35 am on May 17th, 2008 by [Matt]

Somr more info

Speech operated electric wheel chair

A speech recognition system (HM2007) was connected to an Innovation FIRST controller board, and they both where mounted on an old electric wheel chair. Some 12 Sensors were also mounted on the chair for detecting Stairs, obstacles and slops. The man on the chair speaks to a microphone, and the HM2007 recognizes the word and sends its code to the FIRST microcontroller. The C code on the FIRST controller receives the code from the HM2007 and the data from the sensors, and after processing it all together the controller outputs two voltage that tell the chair controller where and how fast to go.

The chair also got an EMIC text-to-speech module. This module vocally speaks out warnings to the man on the chair.

This project has been held at ORT Hermalin (Nathanya Israel) with collaboration of the Israeli Technion Computer Science Department and by Yosi, Ron and Shay, with my guidance

Amnon Demri
2008

Posted at 9:32 am on May 17th, 2008 by Amnon

wow thats really cool and btw he is speaking Hebrew if you are wondering…

Posted at 9:34 am on May 17th, 2008 by pwave

Nice project Amnon. I was originally planning on doing something similar for my university project, but eventually dropped the idea due to cost and objections from my team mates.

My initial plan was to use a cheap laptop (eg. the Asus EEE pc) for the speech recognition, as it would allow for future upgrade and improvement to the speech recognition software. The laptop platform would also allow for greater features, such as a rear and side view camera, audio warning, and easier programming of the control logic. The cost is not much more than the speech recognition board considering the cost of the wheelchair, and the higher power drain should be small, compared with the wheelchair’s drain.

A problem that was considered, is that with a voice recognition system, the speed, accuracy, and granularity of the control would be rather poor. The wheel chair speed would likely need to be kept very low to maintain controllability and safety.

A later concept would be to use a pair of accelerometer and hall effect sensor, one mounted on the wheelchair and the other on the user’s headband. By calculating the angle between the two, the system could determine which way the user is facing. A tilt of the head would move the wheelchair forward, while turning the head will cause the wheelchair to turn in the direction the user is facing. Such a system would provide a natural and proportional control, and I believed that it would allow a higher wheel chair speed to be manageable.

The drawback would be that this could only work if the user have control over his head movement, which many tetraplegics lacks.

I would be interested to know how you have arrived at the combination of hardware used. Also, did you manage to trial the wheelchair with a tetraplegics? I assume that would be it’s target user. Did they find it useful?

Regards.

Cort

Posted at 1:05 pm on May 17th, 2008 by Cort

This is an awesome project.

Reminds me of jokes we were making about speech recognition being included in games.

GOGOGOGOGOGOGOGOGOGO!!!!
NOOOOO STRAFE STRAFE STRAFE!!!!!
GOGOGOGOGO!!!!

Posted at 3:32 pm on May 17th, 2008 by w00tb0t

I can’t find any sort of link for this, but there’s a startup company at UIUC working on something very similar to this, except they’re using sub-vocalization. They showed us their prototype, which did work surprisingly well, at a student IEEE meeting. The transducer is a clotch that velcros around the person’s neck and reads the faint electrical impulses given off when a person “tries” to say something…so actual speech isn’t necessary.

Posted at 6:11 pm on May 17th, 2008 by watts

Why hasn’t anyone thought about making a wheel chair that runs off of the nerve impulses from the brain or damaged legs? We have working examples of people playing video games purely through thought/nerve impulses via a helmet they wear…why not use that in a wheel chair? It was an idea I wanted to go with in my EE project, but instead went another direction.

Posted at 12:31 am on May 18th, 2008 by TicTac

Cort and No6 Hi

You have asked me ” I would be interested to know how you have arrived at the combination of hardware used” ?

I am not sure that I understand your question .

And for your seconded question :

did you manage to trial the wheelchair with a tetraplegics?
We Didn’t try this wheelchair on any other then the project team , me , and the collage principal .the hm2007 (or the way we use it) is not very reliable ‘ so we didn’t take the chance with tetraplegics.

No6 wrote some thing interesting so I Google it and I found :
http://www.think-a-move.com/scow.html

I think that the microphone that No6 talk about is a “neck mic” ‘ this is the same microphone that spatial forces use (Like the navy seals) and it allow you to talk in a very noisy environment.

You all welcome to look some other projects that I did on youtube :

http://www.youtube.com/user/youwillcallmede

I got a lot of other projects that I didn’t upload yet.
And No home page yet.

Amnon

Posted at 1:32 pm on May 18th, 2008 by Amnon

I foresee one potential vulnerability…

http://www.engadget.com/2007/02/01/remote-exploit-of-vista-speech-reveals-fatal-flaw/

User pulls up to the corner.

Bystander: So I was talking to my friend and told him to go forward!

User’s chair drives out into the intersection. User is hit by a truck.

However, the easiest fix for this would be to use one of those vibration sensing mics that attach to your throat. afaik, that should eliminate the threat, or at least mitigate it to someone who really wants you dead, not just the casual killer/accidental occurrence.

(p.s. if you are stupid enough to try and kill someone like this, or in any other way, I assume no responsibility for your dummerkopf actions.)

Posted at 10:57 am on May 19th, 2008 by srilyk