Question about Arduino

I have been planning a purchase of an Arduino after building several robots built a round the Propeller, BasicStampIIpx24, Basic Stamp IIpe,BasicStampII. I found that the fastiest cpu with eight cores was not compatible enough with regular code for my peanut brain to interpret. When I used simply a BasicStampII, all of the things I wanted to do was realized one at a time.

Problem was, when I tried to do all of these things at once, while the machine did them, it was so sluggish. This could have been my lack of education on coding. Except that over 20 years ago, I created many vertical markets coded in DbaseII and Basic.

I want a machine that can be all that it can be. I want it to have artificial intelligence and emotions built in. I know that this is not an easy task. But, at least I will not run out of inputs on the Mega unit, I trust. I just hope that it runs faster than the Basic StampII and that C language has a 1:1 ratio so it should run fast as ASM Language. I am not a C programmer, but I have several books on it and I dabble.

I would just like to have your input. I plan to order a Mega Unit with a LARS protoboard. I am on a tight budget having spent all my experimental funds on the other projects. I will say that I learned a lot and it was quite worth it. The People at Parallax is second to none.

I want my robot to be round and powerful, and very smart!



I don't get what you are trying to do?.

The mega or any Ardunio should be fast enough for most projects, but no way a super computer. ( it is what it is a low power computer used in automotive industry eg for the engine management system ).

I think you can make a robot as smart as you are a programmer. Artificial intelligence!!

I think you can make a robot head toward to light etc, if you call that artificial intelligence. You can get single celled creature like the British “Member of Parliament” do that, but they head towards any form of money.

You can get single celled creature like the British [ch8220]Member of Parliament[ch8221] do that, but they head towards any form of money.


OP: I say go for it!

Ok, Peter, I will try to explain. I don't have Everything in front of me, but i will try to give you the just of the situation.'

I have always wanted since I was a little boy a robot that demonstrated Artificial Intelligence.

years ago, I read two books, one was How to build your own self programming robot. The other was how to build your own Robot pet. Both books went far beyond things people are generally doing with their projects. They generally stop once it is an obstical avoiding robot and they are satisfied with that.

I want my robot to exhibit other factors beyond obstical avoidance.

Using RobotBasic and the book Robot Programmer's Bonanza, and a serial wireless communications, you can create a robot somewhat similar to these robots. Chapter 14-17 explains and they have already written a program that simulates in realtime this situation. I do not have serial communications or the money to buy it, but Xbee is in my future, I hope.

Now, I wish my robot to: 1a. Check his environment, UltraSonic, IR, switches, sensors, etc. 1. start off with generating a random number 2. Use the random number to select a task of about one out of 20 tasks. 3. check to see if confidence level is high or low for this task. Example :If he ran into a wall 5 or 10 times, then the confidence level will be low. But if he has had a clear path in the past in this direction, the confidence level will be high. 4. If confidence is High then do the task. 5. If not, goto 1a. 6. Task will include but not limited to: finding his charger. getting out of a jam or stucked situation. Talking or Singing or playing music. Forward, Backward, Circling, Right, Left, etc. avoiding light phobe or going to light phile. Animating his face when he shows different emotions such as: Happy, Sad, Hungry, Angry, Curious, Surprise, Scared, bored,etc. Responding to Pain or Pleasure. Example: Pain=battery low, Pleasure= Battery Charged, etc. Not actually having a human emotion, but demonstrating as if he did. Being able to learn from his mistakes and adjust his programming accordingly. Example: map the room takes hours, but after it is mapped he knows where things are until you move them, then he uses his sensors to correct the confidence level.

I say he, but this could be a she as well.

Be able to recognize a certain human being by color or maybe the tone of voice. Have Ultrasonic, IR, Photocells, Audio, Video inputs as well.

These are only a few of many possiblities. It sounds like a lot, but what if we store each item as a single bit in a word for the environment? Each Bit would be a part of the Environment. It should not take many words to store it. I am curious about how long it would take to calculate. I am not a professional programmer anymore.

wow, that is complex !!

I would say that voice recognition , video pattern matching are beyond what a ardunio could do.

From what I've seen you would need a super computer for it. On the other hand, I think the sony aibo is that intelligent. But how many millions do you think they have spent getting it right.

But, if I could just get it to remember where it has been and do some of the AI, I could add a little at a time. But, that wasn't the whole scenario. But, I guess you understood what I was talking about.

Build a maze solver, there are a lot of people building robots for that. And that should be simple too, which should help you to think how to program IA into a your super bot.

Most of them simply follow the black tape on the ground. I want the robot to experience the situation.

Thanks for the advice, though. :)

If you want to program gigabytes of A.I and keep your hardware overhead to a minimum, and your o.k with it operating indoors at a range of around 100m - then check out the wifi shield

I don't think anyone's done a wifi-bot yet, and I think there's a few methods of extending the range of wifi so you could have it running around your neighborhood armed with the intelligence and capacity of your host computer = )

The wifi shield is new though and still having tweaks made to it, but by all accounts its working to its specification.

i have a wifibot driving around my house...

with a WRT54G v2.2 12v battery's 2x 1.6aAh a ipcam for coordination and a arduino + ethernet shield for controll with head lights a dukes of hazard horn (a cheap hacked mp3 player with one mp3 on it :P) all placed on a RC toy car!

it can run for 45 minutes...

i would love to built a recharge/docking station :P i have 13x 12v 1.6Ah battery's :P but it would still be nice if the robot would sense low power and return to the docking station!

a dukes of hazard horn (a cheap hacked mp3 player with one mp3 on it )

Nice! I like it!

Your goal can be broken into four big parts...

  • Detect low batteries
  • Locate the docking station
  • Connect to the docking station
  • Correctly / safely recharge the batteries

Which part would you like to discuss first?