Many Questions For Large Project

[mmabigshow]

Greetings,
 all my life I have had a bug for robotics. Now, with a family of 5, we are moving into a new gorgeous house where I will have a workshop for the first time in my life. I am eager to begin a large project.

I am looking to build something like a Wall-E or Johnny 5 type of bot.

Questions:
#1. What should I include in my workshop in terms of tools of the trade that I can't build without?

#2. Mobility: Should I focus on a tank track system, 3 wheel base or 4 legs/6 legs and why?

#3. What would it cost to have my parts custom built over buying various kits? I saw one kit for a track base that costs between $4,000 and $5,000. I don't want to spend that much on a base. What are my alternatives?

#4. My goal is to have a mobile bot that can interact with the kids someday. Something I can upgrade down the road as needed and is able to recognize speech and speak itself. There are a lot of speech recgnition packages out there as well as some AIM Bot software you can run that lets the AI learn discussion, Q&A, Jokes etc, you type questions and you can predeterme responses for it or let it learn on it's own. Might anyone know of any way to use the two together with a voice synth?

#5. Are there any shops that do custom work by design? In other words, a place I can email a plan for a part and have them make it for a cost?

I know the original creator of Johnny 5 was considering selling plans and schematics for people to build their own unit but I haven't found whether or not that was ever offered. Anyone see this before? I am also following the Wall-E builders group as a lurker to see how they are doing things over there.

I know many of you will want to tell me to start with small bot kits and such to get warmed up and I've already done that. I'd honestly rather spend 10 years working on this large bot learning as I go than tinkering further with small kits and such.

I'll leave this here but I'm sure I will have more questions later.

I know how knowledgeable many of you are on this forum and I'm very happy to have found it. Well done!
I look forward to any and all responses.

Thank you.

Jason

[Trumpkin]

#1. What should I include in my workshop in terms of tools of the trade that I can't build without?

You will definently want a good soldering station, third hand tool, and solder sucker (and solder, of course). Also, you might want to get a lathe, I don't have one, but it might be handy for building a bigger robot. You might also want a dremel?

#2. Mobility: Should I focus on a tank track system, 3 wheel base or 4 legs/6 legs and why?

You  should probably go with 4 wheels or 3 wheels unless you want it to go up and downstairs. trying to make a 4 or 6 legged robot could get really frustrating and expensive.

#4. My goal is to have a mobile bot that can interact with the kids someday. Something I can upgrade down the road as needed and is able to recognize speech and speak itself. There are a lot of speech recgnition packages out there as well as some AIM Bot software you can run that lets the AI learn discussion, Q&A, Jokes etc, you type questions and you can predeterme responses for it or let it learn on it's own. Might anyone know of any way to use the two together with a voice synth?

That is probably a good question for airman00!

[R.O.B. 2.0]

#3  Kits may seem expensive, but if you got the same parts in the kit custom machined... it would cost a whole lot more. Kits are also good because the manufacturers know that the kit works or they wouldn't sell it(usually). Custom machine parts are good though if you can't find a kit with the parts you need.

[mmabigshow]

Thank you very much for your responses so far.

[bukowski]

#1) Find out what your budget is and work from there. Get some test equipment. If you are short on funds, just get a good multimeter, if not get a multimeter, a o-scope, and a variable power supply... the list goes on. Along with what Trumpkin said (all good suggestions except for the lathe... If you were going to get any machine equipment I know I would rather have a small mill) get an electronics toolkit (like exelite).

#2 & #4) Depends on your experience, and what you want to do with it. If you are good mechanically, then make a chassis for your bot that you will be using for the next few years.(Imagine your dream chassis. If you can answer the majority of the questions/problems in designing/building it go for it! If not, start downgrading until you can. Also, multiply the number of questions/problems that you have by 10 or so... inevitably you will run across problems that you didnt anticipate [just like any other project]).

#3) Depends on the part.

#5) Yes, check your phone book for small machine shops. Bear in mind that this will be expensive. A lot of the cost in machining parts is set up time, so for one-off or short run parts, you will pay through the nose.

[pomprocker]

Desktop CNC machine and drill press, coping saws, all that shop stuff for building the chassis.



Oh, and you may want to track down Newton Crosby also, as he invented Johnny 5 over at Nova Laboratories, in Damon, Washington

[img width= height=]http://www.fast-rewind.com/shortcircuit2.jpg[/img]

[Trumpkin]

oops i meant mill instead of lathe.  

[airman00]

#4. My goal is to have a mobile bot that can interact with the kids someday. Something I can upgrade down the road as needed and is able to recognize speech and speak itself. There are a lot of speech recgnition packages out there as well as some AIM Bot software you can run that lets the AI learn discussion, Q&A, Jokes etc, you type questions and you can predeterme responses for it or let it learn on it's own. Might anyone know of any way to use the two together with a voice synth?

That is probably a good question for airman00!

most people don't realize how difficult it is for a robot to learn discussion. I am working on something like this , but its pretty damn difficult.  Email me and it'll be easier to talk through there, i've discussed this soo much here on the forum. Search the forum using the search bar . Then after that either email me or post a topic.


Anyways I'm writing a tutorial on my butler robot which does some of this

[Admin]

My #1 advice is to start on a simpler project first. Learn the basics before learning the advanced stuff. You can always reuse parts/equipment for the more advanced project later. Make your mistakes on cheap stuff first

(and I don't mean kits, I mean something like my $50 Robot)

#1. What should I include in my workshop in terms of tools of the trade that I can't build without?

Whats your budget?
oscilloscope, rework station, drill press, band saw, mill, vice . . . those alone will probably cost you $5k-$10k total . . . if you are settling down forever, get some good quality equipment and research it heavily first.

#2. Mobility: Should I focus on a tank track system, 3 wheel base or 4 legs/6 legs and why?

wheels . . . thousands of times easier/cheaper . . .

#3. What would it cost to have my parts custom built over buying various kits? I saw one kit for a track base that costs between $4,000 and $5,000. I don't want to spend that much on a base. What are my alternatives?

depends entirely on what you want made . . . doing something custom costs you time, and the cost of equipment, but you learn more and have more pride in the work . . . being a mechanical engineer, I make everything on my own

#5. Are there any shops that do custom work by design? In other words, a place I can email a plan for a part and have them make it for a cost?

http://www.societyofrobots.com/robot_parts_list_manufacturing.shtml

[coldkryten]

#1. What should I include in my workshop in terms of tools of the trade that I can't build without?
#3. What would it cost to have my parts custom built over buying various kits? I saw one kit for a track base that costs between $4,000 and $5,000. I don't want to spend that much on a base. What are my alternatives?
#5. Are there any shops that do custom work by design? In other words, a place I can email a plan for a part and have them make it for a cost?

A variable power supply is remarkably useful.  Good soldering iron with changeable tips (look at Weller or Hakko.)  If you want to do a lot of your own electronics, a digital oscilloscope is well worth the extra money over an analog scope. 

Second the drill press, band saw, vice, dremel, etc...  A cheap router can be remarkably useful.

Having shops make custom parts for you is expensive, generally expect to pay a few hundred dollars for just about anything.  When it doesn't work and you need a second version, it's going to cost you the same again.  Sometimes you can find a shop/machinist who will do discount work for a hobbyist, it's still going to add up - kits will definitely be cheaper than this. 

From what you've described, learning to machine would be a good option.  A desktop lathe is somewhat useful and can be gotten quite cheaply.  Desktop mills are a bit more expensive, and often with a little creativity you can get similar results with the drill press/band saw.  You want want to think about a small CNC mill or router.  This will let you build all sorts of crazy things and will grow more useful the more you do.  It will certainly pay for itself in a few years if you plan on sending out lots of custom parts to shops.

#2. Mobility: Should I focus on a tank track system, 3 wheel base or 4 legs/6 legs and why?

2 wheel drive with one or two casters is BY FAR the easiest solution.  Limited to fairly flat and even surfaces and not as cool.

Tank treads are a bit more challenging but you can drive over uneven ground (i.e. outdoors) which is a big plus.

Build a 6 legged walking robot and get it to move around before even considering something large with any number of legs.

#4. My goal is to have a mobile bot that can interact with the kids someday. Something I can upgrade down the road as needed and is able to recognize speech and speak itself. There are a lot of speech recgnition packages out there as well as some AIM Bot software you can run that lets the AI learn discussion, Q&A, Jokes etc, you type questions and you can predeterme responses for it or let it learn on it's own. Might anyone know of any way to use the two together with a voice synth?

Doing anything like this will require a lot of programming knowledge and time.  There are several speech recognition and voice synthesis libraries for C/C++ and Java.  Look up speech recognition in Vista on youtube to get an idea of what you can expect, and think of how much work Microsoft can afford to put into something like that.

[JesseWelling]

#1. What should I include in my workshop in terms of tools of the trade that I can't build without?

Start with soldering iron and dremel and multimeter for sure. A drill press, band saw, and lath, in that order if you have the cash...

#2. Mobility: Should I focus on a tank track system, 3 wheel base or 4 legs/6 legs and why?

Track or Wheels are easiest to get going on. If you build the torso/arms/head/scorpion-tail so that it can bolt on to anything, you can switch to legs when you get more experience working with controlling servos and can handle more complexity.

#3. What would it cost to have my parts custom built over buying various kits? I saw one kit for a track base that costs between $4,000 and $5,000. I don't want to spend that much on a base. What are my alternatives?

Lynxmotion has some servo erector sets and I'm sure they would custom build you a solution, but I know that this may also fit your bill, and it would seem to be pretty cost effective for what you want....

#4. My goal is to have a mobile bot that can interact with the kids someday. Something I can upgrade down the road as needed and is able to recognize speech and speak itself. There are a lot of speech recognition packages out there as well as some AIM Bot software you can run that lets the AI learn discussion, Q&A, Jokes etc, you type questions and you can predetermine responses for it or let it learn on it's own. Might anyone know of any way to use the two together with a voice synth?

Good luck with all that. It's not in my realm of interest at the moment so I have no clue  But you will most likely want to go with a layered architecture then. Like a microcontroller that does all the low level mechanical control and a high level AI on a PC connected through Wifi, or an on-board SBC talking to the microcontroller.

#5. Are there any shops that do custom work by design? In other words, a place I can email a plan for a part and have them make it for a cost?

Polulu does some custom laser-cutting. And you might check out Phenostream. They look pretty cost effective as well.

I know many of you will want to tell me to start with small bot kits and such to get warmed up and I've already done that. I'd honestly rather spend 10 years working on this large bot learning as I go than tinkering further with small kits and such.

I'm not going to say start small, I'm just going to say make modules that are loosely coupled so you can work on sections that can be re-factored or replaced completely at a latter date. Also, you can jump around to different modules to avoid burnout on any one problem... And read books on theory if you get burnt out on coding, it seems like a waste but it's more productive than doing something unrelated if you really do need a break.

  .... and may the force be with you.

[Cotowar]

#1. What should I include in my workshop in terms of tools of the trade that I can't build without?

Coming from a background of significant mechanical construction, I would highly recommend a band saw, drill press, and compound mider box over anything else. These three alone are an entire shop if you know how to properly use them. As for other machines, a router is nice to have, a welder(if you can weld), a dremel, and a few hand tools are about it. Depending on the materials you work with, other tools might be nice, but are not necessary.  (the other machines I included are based on personal preference).

#2. Mobility: Should I focus on a tank track system, 3 wheel base or 4 legs/6 legs and why?

Tank tracks are hard, and feet are even harder. Stick with the wheels unless you know exactly what you're doing.

#3. What would it cost to have my parts custom built over buying various kits? I saw one kit for a track base that costs between $4,000 and $5,000. I don't want to spend that much on a base. What are my alternatives?

To be honest, depending on your skill with machines, and your ability to visualize moving parts, you would be best off to make custom stuff yourself. Otherwise I would recommend buying kits and scrounging for parts.

#4. My goal is to have a mobile bot that can interact with the kids someday. Something I can upgrade down the road as needed and is able to recognize speech and speak itself. There are a lot of speech recgnition packages out there as well as some AIM Bot software you can run that lets the AI learn discussion, Q&A, Jokes etc, you type questions and you can predeterme responses for it or let it learn on it's own. Might anyone know of any way to use the two together with a voice synth?

I would just research the hell out of this question, I personally don't know enough about the topic to help you, but am also planning on making something like you are.

#5. Are there any shops that do custom work by design? In other words, a place I can email a plan for a part and have them make it for a cost?

Any metalworking shop should be able to do your metal items, and woodworking shops can do your wood/plastic items. I personally would recommend making everything yourself, but for complex stuff, just check the yellow pages.

I know many of you will want to tell me to start with small bot kits and such to get warmed up and I've already done that. I'd honestly rather spend 10 years working on this large bot learning as I go than tinkering further with small kits and such.

Thats how I work too man. I'm trying to jump right in on what will eventually be a pretty capable robot. If you need any help on mechanical stuff, let me know. I'm sure I can talk you through how to make something work. For the electrical stuff, I'm no good, and i MIGHT be able to help with programming, otherwise, its up to the other fine people on this forum to answer your questions.

Hope my answers help you out a little. Cheers.