An robot arm as schoolproject, is that a good idea?

[Fredrik Andersson]

Hello everyone!

I'm pretty new to robotics, the only robot i've done is an 50$-robot sort of robot. I have the expirience of two months of reading (could have created some more of a robot, just that i am waiting for parts that are never coming ). Im a former game programmer (if you like to entitle yourself with your main intrest that is) so i am atleast fit at the programming part.

Ok, here's what i want to do. I want to build a robot as school project, an really advanced one. I got an really cool idea that may actually be of intrest for robotics laboratories. Of course it's only idea now, so i will need to have a prototype to show some intrest, right? Now it may have been created already or is kind of an bad idea,but that is why im asking. Im still concerned of trying though.

Right, the idea... the concept is to create an mobile robot arm, that could act like an third arm or replacement for one. By this im not meaning some kind of an prosthetic, connected to the nerves and stuff like that. Im meaning more like an autonomous robot that responds on voice commands. The arm coould be commanded to do all kinds of things like picking up an object and drop it somewhere else, picking up a cup of coffey, and pouring it into the owners mouth (yikes! if the coffey is hot, the arm should have temperature sensors and refuse to pour if it's over an maximum temperature or something like that ) turning pages on a book... the list goes on.

Now im not aiming for the robot to be practical, not my version. Im aiming for it to be cool and make an great appearance so that it can maybe show some intrest at some company or laboratory. The only practical? thing i want it to be able to do, is to be able to pour coffey in my mouth!

While i need to know if there may be even a slight intrest in this thing and if i will be able to do it i would also like to know if i can get funding for it  somewhere. Im guessing that the advaceness of it will be linear to the funding for it. Should i talk to my principal and maybe get some funding from the school? If so, would my chances be better if i prepared a little presentation telling him about my higher purpose of the robot, and how it will draw attention to the school with such an creation? Also, Could i get some grant from the state (Sweden) and if so, how do i do that? Should i send an documentation on the project and consept? I should propably ask this to my teachers and principal instead, but i guess there is something you have to say about it. Bring it on!

Have a great day/
Fredrik Andersson

[airman00]

This is some really advanced stuff! the $50 robot compared to this is like a paperclip compared to a toaster.

If you really think you know how to build this arm then buy a cheap toy robot arm and hack it. Try the OWI Robot Arm.

As for the whole voice command thing, that seems kind  of pointless to me.Example: "robot grap cup" = has to find a cup and grab it. Then "robot bring cup to my mouth"= bring cup upwards to a set location without spilling. Now the tricky part... drinking. How would it know where your mouth is, or if it is open or closed?
DIFFICULT.

For voice commands you could use the HM2008 chip or just use a PC with voice recognition software. Check my blog for details about voice commands.

if you decide to  start this project please post pics of it on the forum when you're done. This seems like a really cool project.
,Eric

[Fredrik Andersson]

Hi, and thanks for the reply.

As for the cup, i was just joking. My prototype doesn't need to do that, but it would be something for a real production of this kind of thing. Mine just need to look cool ^^

How about this chip? http://www.imagesco.com/speech/speech-recognition-technology.html I don't need more that 40 different commands. Objects to pick up could just be recognized through color and don't need to be geometrically advanced.

However, i forgot to tell you some about the "higher purpose" of this project. If some professionalls can creat an similar thing, it could get really advanced, advanced enough to help someone that are lame but can still speak, or have lost an arm or two. Sure, there are far better solutions, like connecting the robot arm to the nerv system or muscles, but this arm would be much cheaper and wouldn't need an operation.

[airman00]

oops I meant that chip. HM2008 instead of HM2007. lol 
Yes that kit fits your requirements. There are two kits to bu there. The speech recognition kit AND the speech interface kit. The interface kit gives you a logic signal on a certain pin whenever the corresponding word is detected. Just connect the logic signal to a micro controller and program it to react accordingly.
The total would be ~$200. One problem with that system is that the recognized phrases have to be short. You cannot have it recognize full sentences ( maybe if you have it recognize word after word...)

A cheaper and easier approach if you have a laptop you are willing to "donate" to your robot arm project, is to buy a Phidgets Interface Kit and have the computer listen  for select words or sentences and then send a signal through the phidgets to your robot arm. The problem with that is that it requires a laptop which takes up lot of space compared to the HM2007 circuit which is very small. Check on my blog (erobots.blogspot.com) for more info on this method.

If you are thinking of an arm that is small just go with the speech kit
If you have any questions just ask...

[HDL_CinC_Dragon]

Dont mean to just kinda slap a penny down on this 20 dollar convo but havnt they worked out a way to receive brain signals withOUT electrodes directly in the brain/nervous system? Some kind of helmet thing that picks up on minute EMFs from certain parts of the brain or something of that nature?

[airman00]

Really, that sounds really cool!

Do you know if its available to the public for purchase. Also, how can you tell which brain signals correspond to which action?

I want one of those 

[HDL_CinC_Dragon]

lol like I said im not too sure. I could be mistaken. I dont think the human brain can even generate such an EMF... who knows. Maybe I dreamed it hahaha

Thought I saw a video of it online somewhere.... eh well.

[awally88]

I've heard about something like this, I'm not 100% sure how it worked, it was on television a while ago but they made an actual arm for a person which was connected to their brain.

From what I remember someone from the US army got their arm blown off in combat or something similar and then a new arm was made for them.  Allegedly she had similar control over this to a normal arm!!

I'm not 100% sure if this is exactly true but this really is an amazing feat!

Also Fredrick this sounds like an awesome project, good luck with it!

[Fredrik Andersson]

Yeah i've seen an article about that helmet, i think it was in connection to gaming. It seems cool though, and i would really like to get my hands on it and try it out for the arm. Im guessing it's pretty expensive though.

I need to ask again, do you think that this idea could acually sell? It will be a LOT cheaper than an "real" robotic prosthetic and money is probably an issue for any disabled person. It may not be very practical and the solution causes slow action, but i thinkthose people may still find it as an freedom beeing able to do some things on their own that they else wouldn't be able to. And, this arm means no operation and could be used right away after the purchase...

[HDL_CinC_Dragon]

Well too be perfectly honest with you, I dont think it would sell very much at all

I mean especially for someone who was in combat who lost an arm, and then to go from 10 minutes worth of things happen in 2 seconds to having to talk to their arm to maybe get it to do what they want is just something I dont think would happen. Personally Id rather nave only one arm then have to wait like a whole minute to just pick up a can or something... its a great project idea but for an actual practical base for a cheaper prosthetic, not very likely

HOWEVER, that does not mean it doesnt have a possible future, you could make a robotic arm or set of arms for easy and economic use in workshops or something. Have your robotic arm hold things for you or grab things for you. put an arm on a track that runs from your desk to your large warehouse style shelving, grab a box of resistors that you asked it to and have it bring it to your desk for you.

"ShopHelper 1, retrieve box 23 from shelf 1. ShopHelper 1, retrieve box 17 from shelf 2. Go"
"ShopHelper 2, retrieve box 36 from shelf 4. Go"

Then the 2 robot arms would travel down their tracks, grab the boxes they were told to, and bring them back to you all-the-while you continue to solder your boards and such without having to get up and put the soldering on pause.

Just a thought.

[HDL_CinC_Dragon]

just came across this on the PopSci website. Saw it in the newsletter they send out
http://www.popsci.com/popsci/technology/5dfc3bd0efe84110vgnvcm1000004eecbccdrcrd.html

[Fredrik Andersson]

Yeah, i never though the prostethic tech has come that far. I did a search for it and it seems there are already much better tools, even without any electronics involved. And the myoelectronic technique is far better for getting commands than voice commands ^^ they simply monitor the small electrical signals at the stumps end and use them for movement...

However, i still want to do something involving an robot arm. Maybe something like your ShopHelper, and mobile and higly dynamic robot arm. How about an more advanced AI compared to all preprogrammed robot arms? It would be neat to adapt the thinking of the AI-talk that is posted on the videos forum so the robot learns all situations. You can for example teach it an position by making associate its position with the particular word you are commanding it, or even that there should be an object at that position. If there is not it will somehow react on that. That way you can teach your own "desk helper" to your own needs ^^

I really do want to do something new, the idea that you actually invent something as a school project is so darn cool. Help me out here with my brainstorming

[HDL_CinC_Dragon]

Hmm that would be pretty darn tough. You would have to show an object to the robot, tell it what it is and show it where it would normally be located and such. Then if it didnt find it in its normal spot it would have to search for it which could take a long longer for a robot to do than a human and such... I dont mean to shoot down your idea but I think you should start with something smaller.

Perhaps mount an arm on a small vehicle chassis and have a series of shapes and colors. Like have some red cubes, some green cylinders, and things of that nature that it can easily recognize and have it stack them or something. Not very practical but for a school project that would be great! Not to mention you could always continue to work on it afterwards and upgrade its hardware and software and add stuff to it and stuff like that.

[paulstreats]

I need to ask again, do you think that this idea could acually sell? It will be a LOT cheaper than an "real" robotic prosthetic and money is probably an issue for any disabled person. It may not be very practical and the solution causes slow action, but i thinkthose people may still find it as an freedom beeing able to do some things on their own that they else wouldn't be able to. And, this arm means no operation and could be used right away after the purchase...

I think that if you make the arm as planned, even with a speech recognition, you would have to build a set of functions anyway that occur when a specific command is spoken.
If you already have this function set, you could then attach any other input device, be it speech, this helmet thing or buttons that interface to your function set.
The set of functions is the most important thing because these are what the speech hardware will interface to like if you say "forearm forwards" this would trigger a function that moves the forearm forwards, so you could also use just a button which when pressed would trigger the forearm forwards function and any other type of device. The input device is therefore interchangeable.

[tigy888]

i also made a robot arm using cardboard but i still haven't added the components yet...