Hi. I am evaluating my ability to run a session of the new Boy Scouts "Robotics" merit badge. Rather than making use of Mindstorms or $100 kits from BSA (first link), I am considering having the kids make the simplest of intelligent agents based on Arduino.
I saw the tank project described below in second link and am going to try to duplicate that. It looks like it can be done for around $65.00. The tank treads seem cheap enough but one thing that worries me is the author's description of how hard it was to get the tracks physically onto the wheels. That is actually one thing I'm not interested in teaching: it will be hard enough to get programming & wiring down. In my extensive googling of the topic :* I've seen other wheels, but they seemed to start at $60-90 themselves. If anyone has tips on simple cheap wheel kits, that would be greatly appreciated. thanks!
Must it be a kit? If not, a great way to have a nearly ready-to-run robot chassis is to go to Goodwill or other thrift stores (like Savers) and buy old toy R/C cars; don't spend more than $5.00 on the car. While it won't have tank (differential) steering - it will be cheap. You'll have to do some "hacking" once you open it up, to identify how to control the motor and the steering mechanisms (some use a motor for steering, others use electromagnets). The best way to tackle this is to trace back from the motor to the control circuits (mosfet/transistors), then find where those are activated. Then, you have to cut traces and such to leave power to the motor control systems (h-bridge or whatnot), but not power the rest of the circuit. Then connect lines from the Arduino (digital outputs) to the control inputs on the PCB, and hook the ground of the Arduino to the PCBs ground.
I have been working on developing a low-cost vehicle-type robot kit from products I have access to in China. Can you tell us more about the most-important attributes for such a kit?
Maybe??
Final shipped cost
Simple to mechanically assemble or pre-assembled (I'd prefer kids assemble)
Arduino based, so easy to program, soon can use drag-drop programming like http://www.modk.it/ (under development)
I have done Lego Mindstorms with kids for years, and also simple servo-motor robots with Dios and now Arduino.. Two servos with wheels, a Nano with a small breadboard, some chopsticks and tape and a 4-AA battery clip and it's moving But I'd like to maybe see something that is more packaged for troops that don't have more-experienced leaders for this badge.
My Son is a Scoutmaster in Vermont and he was looking at maybe doing this badge; I'll ask about collaboration on that...
pterandon:
But cr0sh, the one thing is that I may need to duplicate this for five to fifteen boy's projects. But that's a good idea to look at cheap R/C parts.
That's understandable, but maybe you could turn it into a learning experience for everyone involved.
First, learn how to do it yourself - to the point where you understand what you are looking for, how to troubleshoot and trace the circuit to find the proper points of connection and breaks, etc. Come up with a "how-to" list. In the scenario that the reciever or h-bridge control system is shot, learn how to hook up your own h-bridge to control the device (an L298 or similar would be perfect for the smaller cars, most likely).
Then teach the kids how to do it themselves - the troubleshooting, the tracing, the logical steps in reasoning and thinking to go from "junk" to "robot"...
This kind of knowledge is something that they can apply throughout their lives; essentially, you would be teaching them how to "hack" the stuff they own - to re-purpose things thought of as junk or old into something new and useful, as well as educational.
IMHO, such knowledge is far more valuable and better to impart than comparatively simple kit building...
Terry, the first link I shared has, on the right margin, the literal requirements for the merit badge. I think the key features for designing a kit are:
a. With your counselor’s approval, choose a task for the robot or robotic
subsystem that you plan to build. Include sensor feedback and programming in
the task. Document this information in your robot engineering notebook.
b. Design your robot. The robot design should use sensors and programming
and have at least 2 degrees of freedom. Document the design in your robot
engineering notebook using drawings and a written description.
Now I would suppose that "going straight" is 1 degree of freedom, and "turning" is another, right? The cool thing is that the sensor feedback is unspecified. Could give the kids options if including more than one sensor negligibly affects the overall cost.
As far as cost, I have no bankable number. Somehow I was thinking that the BSA-supplied kits at $100 were prohibitively high. And I'll say it: too cutesy! This could be gotten around I suppose by having them re-used, but that has too many problems. I was comfortable with saying that the price of (kits for) the badge would be $65, but I haven't gotten any feedback yet as to how reasonable that is-- I'm new in this Troop. And by kits I'm thinking a bag of parts that I put together. I'd like to avoid soldering.
Yeah, I'm in the "checking out if I can do it first" stage. Or, "buying the parts" stage. I can program in a few archaic languages and I'd tinkered enough previously with Arduino to make a light blink, but then my young son got bored and we didn't pick up again. He's a little bit older now and with a look at a Boy Scout meeting insisted I volunteer to chair this badge committee.
The "Degrees of Freedom" are covered by Translation (movement in the plane) and Rotation of a typical vehicle. (See Wikipedia...)
There has been discussion on the Boy's Life site about "Design your Robot".. I think the conclusion is that starting with a basic platform and designing the action of the robot, it's movements, one or more sensors and feedback in which sensors affect behavior through Scout-written code is significant 'design'. I would agree.
It's 9AM in USA and I'll try to check in with My Son The Scoutmaster 8)
I like the idea that you use an open-source hardware and software approach. Nothing wrong with Parallax and etc. but they are expensive and proprietary. Arduino is very open-ended. I bet the BSA Tekkie community is large enough to build up a nice collaboration.
Send me an email and I'll send you the ArdBot articles which may be a good basic direction...