A community project for the summer perhaps?

[ericthered]

Modular may sound hard, but it isn't.
Modular is the way to go with multiple people working on one project, it's just pure logic...

And with a forum filled with robotics enthousiasts helping, the project may be a bit hard.

But what about sensors, I mean otherwise we would be doing what everyone else is doing.  Plus modular robots are way too slow I mean that MIT video was speed up 20X. I like the ed209 Idea better. We should make It big enough to fire twin airsoft guns at targets. Some of us can develope walking while other shooting.What company wouldn't sponser a walking machine of death?

[Rebelgium]

Modular may sound hard, but it isn't.
Modular is the way to go with multiple people working on one project, it's just pure logic...

And with a forum filled with robotics enthousiasts helping, the project may be a bit hard.

But what about sensors, I mean otherwise we would be doing what everyone else is doing.  Plus modular robots are way too slow I mean that MIT video was speed up 20X. I like the ed209 Idea better. We should make It big enough to fire twin airsoft guns at targets. Some of us can develope walking while other shooting.What company wouldn't sponser a walking machine of death?

modular doesn't mean a block type robot like in the video. Modular means you can take every part of the robot apart. And connect the parts (like the motor controller print, and the brains, ...are all seperate parts/modules)using only a few wires. I²C preferably.

And I like the idea of a ed209 walker,  You have alot of different stuff going on, so there's enough to work on.

[ericthered]

Your thinking of making a robot like this then? 
[youtube]4oSavAHf0dg[/youtube]
does soundlike fun but I think there should be a couple of different modules for the project like a module with eyes to see and reconize objects. A gripper modual to pick things up, regular modulars to connect and move, and modules with wheels attached to the motors instead just a bunch of connector. What should we use for communication then? ed209 would make a good next project then

[Asellith]

Noone has one the google X prise thing yet. Any rocket scientists in the room? All we need to do is get a robot to the moon and drive it 500 feet with a live video feed in HD quality. So some sponsors and our combined robotics knowledge we could do it just need some rocket scientists to get it to the moon. Of course with something like that we would need a massive amount of coordination and lots of work but would be fun. Specially if we get sponsors to give us free toys to play with.

[Fredrik Andersson]

Noone has one the google X prise thing yet. Any rocket scientists in the room? All we need to do is get a robot to the moon and drive it 500 feet with a live video feed in HD quality. So some sponsors and our combined robotics knowledge we could do it just need some rocket scientists to get it to the moon. Of course with something like that we would need a massive amount of coordination and lots of work but would be fun. Specially if we get sponsors to give us free toys to play with.

Neat, I've dreamed about being involved in a space mission some day. If you guys are serious about it, then so am I! Don't know how much it will take though. The main problem are the rocket. No problem for us building the robot itself, but the rocket, that's the biggest problem. Maybe if we would team up with external rocket scientists.

[Trumpkin]

just connect a ton of bottle rockets to it! LOL 

[Rebelgium]

No problem for us building the robot itself, but the rocket, that's the biggest problem. Maybe if we would team up with external rocket scientists.

Building a robot to got to the moon isn't easy. There are much MUCH more restrictions and demands. (temeratures, rough terrain,...)
Realisticly seen it's impossible for us...
We're still only a bunch of robot hobbyists scattered across the globe...

[Trumpkin]

yep i'm pretty sure that going to the moon would be impossible. I still think we sould do a block design (just not with magnetic plates).

[paulstreats]

maybe if u made enough magnetic blocks, you could build them tall enough to reach the moon

Also im not sure that youre allowed to just send stuff there anyway, i think it has to be built in a sterilized and dust free environment so you dont take any bacterias with you that can be confused for native life

[airman00]

are we going to actually be building this , or not?

[paulstreats]

My thoughts would be to build standard fit modular components, and then ask admin to make a special area of the site so the designs, schematics and programmings could be shown.

Things like:

 Admins atmega board for the $50 robot.
 The same thing but for PIC's
 I2C eeprom extension board
 Servo Controller boards (tell it where you want the servos and it calculates accelleration, Kinematics etc..)
 an intermediary board between the sharp i/r and the main processor, so you processor just gets pure distance measurments

These kinds of modules would be helpful to everybody, and make a great reference library for everybody.

Dont forget that not everybody can make a modular style robot, they require extreme custom machined parts and smd circuit boards to keep their size within reason

[Trumpkin]

I think the first step we should take to build this robot would be to figure out what microcontroller we will be using (AVR, PIC etc.) Will we need more memeory than the Atmega 168? I we don't I say we go with that and the $50 robot microcontroller board. What about making a flying robot? I think we should start a poll to see which idea is most popular

[Rebelgium]

The only way this project can succeed is by having a good leader.
There will need to be someone to divide tasks and follow the progress etc.
It needs to be someone with quite a bit experience on every aspect of building robots and who is willing to put time and effort in this project.

[Trumpkin]

Admin? 

[airman00]

wait we still haven't decided what we are making!

[hazzer123]

I think a leader could help us with the choosing an idea aswell

[cooldog]

notice how admin hassent commented on this topic even though it is 2pg long

[Fredrik Andersson]

wait we still haven't decided what we are making!

However, it really seems like we will be making. Just don't loose this enthusiasm before we decide on the idea.

I'm offering myself as a leader, though I don't know if i have the required skill levels to be it and I can't promise I'll have time enough as I'm studying pretty hard nowadays. If nobody else is offering their help as a leader, I'll take the job.

"an intermediary board between the sharp i/r and the main processor, so you processor just gets pure distance measurements" This sounds like a good idea, and it will absolutely bring value. Perhaps there should be connections for several IR sensor on one of those boards?

[Fredrik Andersson]

notice how admin hassent commented on this topic even though it is 2pg long

I think he is amused of how the community starts to develop on it's own, and decided he would just wait and see where this is going ^^

[hazzer123]

Ill offer myself up for leadership aswell, but i think this project really needs the leadership of an 'Expert Roboticist' e.g. dunk, jon hylands, hgorden or jessewelling. (i think thats all of them)

[airman00]

what will be the functions of this robot?

[cooldog]

Ill offer myself up for leadership aswell, but i think this project really needs the leadership of an 'Expert Roboticist' e.g. dunk, jon hylands, hgorden or jessewelling. (i think thats all of them)

i nominate airman00

[Fredrik Andersson]

Ill offer myself up for leadership aswell, but i think this project really needs the leadership of an 'Expert Roboticist' e.g. dunk, jon hylands, hgorden or jessewelling. (i think thats all of them)

i nominate airman00

I think you mean hazzer123, right? No offense airman00.

[Trumpkin]

dunk seems to have lots of experience i think he should be the leader of this project

[Trumpkin]

I think we should see who wants to be the leader(s) of this project and then have a vote on who those leaders should be.

[airman00]

WE STILL DID NOT DECIDE ON WHAT WE ARE BUILDING PEOPLE!!!     

I understand a robot, but what will the robot look like? What will it do ? etc.

[dunk]

** o wow, 5 posts while i was writing. guess it's still relevant. posting anyway... **

what will be the functions of this robot?

the way i interpret this thread is we are designing a set of modules that can be combined to produce whatever robot you choose to make.
(personally i think the robot built from blocks in the vid above is cool but not really capable of getting beer from the fridge or shocking the cat which are the main uses for robotics in the RealWorld (tm).)

example modules that i think valuable to a project like this are as follows:
  - servo controller.
  - motor controller.
  - I/O module. (able to read inputs from digital or analogue inputs and output to digital or PWM pins.)
  - wireless UART link (for remote control of projects)
  - modified servo + wheel combination (with data on power draw, how fast it travels, how heavy a robot it can move etc.)
  - motor + wheel combination (with data on motor controller specs, how fast it travels, how heavy a robot it can move etc.)
  - mapping processor. (able to take input from I/O modules and output commands to servo or motor modules)

so imagine you want to make a killer 8WD bot with a robotic arm for grabbing beer out of the fridge.
rather than having to design your own components, just build 8 of the motor controller modules, 8 of the motor + wheel combinations, some servos and a servo controller module.
now connect them all to a wireless UART link module and connect the other end of that to your PC.
at this point you find no one has written a nice program to control 8 motors plus all your servos from a PC so that becomes the "module" you decide to design.

the key to this sort of modular approach is coming up with a set of standards which means that all the modules are controlled in the same way.
for motors this is fairly easy. there are currently 3 main standards i can think of: regular DC motors, stepper motors and modified servos.
i would suggest we allow all 3 to be developed as valid modules. (obviously motor controller modules for each would be the first step.)

a harder thing to decide upon would be the communication method between modules.
i have some thoughts on the matter but rather than go into too much detail i suggest we allow UART and i2c communication methods between on board hardware modules
and allow RS232 and USB communication between robot hardware and a remote computer (only needed if your project needs remote communication).

it could be argued that we need to also choose a standard microcontroller but i would argue that we do not.
as long as you can program your microcontroller to play nicely with the other modules i don't think it matters if you use a PIC or AVR or which version of each you use.

what would be important is to come up with a few reference microcontroller designs with working i2c and UART code and circuit diagrams so novices could modify these rather than having to start from scratch.

Ill offer myself up for leadership aswell, but i think this project really needs the leadership of an 'Expert Roboticist' e.g. dunk, jon hylands, hgorden or jessewelling. (i think thats all of them)

flattering but i really doubt if i would have time to lead this.
i am completely up for helping with initial design specifications though.
also expressing opinions. i'm good at that.

i appreciate the need for inter-module communication would put this project out of reach for complete beginners but most of the people enthusiastic about this thread have at least some experience....

thoughts?

dunk.

[ed1380]

i think this project really needs the leadership of an 'Expert Roboticist' e.g. dunk, jon hylands, hgorden or jessewelling. (i think thats all of them)

+1
can't forget admin

[Ro-Bot-X]

Dunk, I take off my hat in front of you. I am also thinking the modular way is the best way to get something standardised. You beat me to this answer as I was still thinking how to explain everything. I don't have anything to add at this moment. I just hope someone else will be making the "code module" cause I suck at programming. (Just make it easy to understand for novices, please.)

As for the leader, it matters not who the person is. The most important thing is who will be making which module?

Edit: let's state the communication specs between the modules so we can start building them!

[cooldog]

we all specialized in certain Fields. dunk with his internet control, airman00 with his butler bot voice recognition. but we are all saying we need one leader and i think that is true try admin for example if he was the leader he would see over the entire project. then we should have group leaders (programming, contruction)

WE STILL DID NOT DECIDE ON WHAT WE ARE BUILDING PEOPLE!!!    

I understand a robot, but what will the robot look like? What will it do ? etc.

i totally agree. i think we should go all terrain. although household robots are really cool and are usefull.

Ill offer myself up for leadership as well, but i think this project really needs the leadership of an 'Expert Roboticist' e.g. dunk, jon hylands, hgorden or jessewelling. (i think thats all of them)

i nominate airman00

I think you mean hazzer123, right? No offense airman00.

haha no i was serious he would make a great team leader or even a group leader