Posts tagged Robotics
Learning a Graphical Programming Language
Jan 26th
LabVIEW robot
Back in my youth, I was known for my study of programming languages. I’ve lost count, but I’m sure I knew a few dozen different languages from ADA to Z++. I learned LISP to support AI grad students and to program my own Emacs functions. I learned SNOBOL before there was Perl. APL was probably the strangest. Writing a cross compiler to develop parallel processing code on a Sun workstation was the most challenging.
Now I’m learning LabVIEW, a graphic programming language built to process data flows. The “code” executes from left to right with program structure and “visual instrument” modules wired together. Input wires are on the left of objects; output wires are on the right. Code such as a while loop does not execute until all of the input data is available, and does not exit until all of the output has been sent.
I haven’t been productive yet, but the clock is ticking. I have to mentor my high school robotics team on how to design and write code asap. Our robot ships on Feb. 23rd. Yes, FedEx shows up to transport the robot (still needs a name!) to a holding facility until our regional competition on March 11th. Everyone has the same length build season, regardless of when their regional event is held.
KickerBot Design
Jan 18th
KickerBot v1
Here is my partially complete robot designed to kick a small ball using rubber bands as the energy source, powered by a rear mounted winch. At our Saturday FRC team meeting, we watched a video of a ball kicking mechanism. I believed it used a rack and pinion device to cock the leg back which then, lacking a lock and trigger device, kicked automatically. I don’t have the lego pieces for rack and pinion (sigh), but a winch should work, using a rubber band to return the winch to the start position.
This is my first design from scratch. I had built the ShooterBot which uses a front mounted piston to shoot small balls out of a magazine rack. The magazine is a plastic piece designed to require a certain amount of force to release the ball through the front hole. The arm of the “piston” is actually stationary, the motor/servo moves the entire ball magazine across the length of the piston arm. I’m sure there is a better technial name for that.
This guy has a “leg” hanging just below the front edge of the NXT brick. As the winch pulls back on the leg, rubber bands (not shown) are stretched. As the leg rotates up, the cocking mechanism (not built yet) will eventually lose grip of the leg and the bands will shoot the leg forward. A different band will then pull the cocking mechanism back into place as the winch unwinds.
This design is currently top heavy and unstable; the frame needs more work. But I didn’t want to build anything else until I got the cocking mechanism completed. But I might move on to building something else now that I see that I can use a motor as a simple winch. I’m more (much more) of an idea guy than a mechanical engineer. I also have a terrible time finishing projects. Hmmm.
Google before breakfast, build a vacuum pump by dinner
Jan 15th
Vacuum pump parts
For our FRC robot, which is playing a soccer-like game, we need a way to possess the soccer ball while maneuvering on the field and over the mid-field bumps. We can’t pick the ball up off the field, and we can only continuously touch the rear 3 inches. We can also touch most of the ball (if we are very close to it) for 2 seconds at a time, every 2 seconds; in theory that is to kick the ball but I suppose we could design top or side grabbers.
The first design the team had was to use a vacuum. We were excited about that when we were thinking of a car or shop vac, but when we realized we had to *build* the vacuum… well, how do you do that?!? The first idea was to reverse a fan, which might work. But my morning Google search turned up this plan to build a vacuum pump. It is always better to have two options, and I think either might work.
Penguin controlled robots
Jan 13th
TI BeagleBoard
Of course, lets not forget that I’m a Linux Architect… I need to get a small linux system embedded into all of my robot designs. There are a mess of tiny single board computers out there, and Linux is the operating system of choice for both hobbyist and professional.
The smallest one I’ve seen was actually embedded in a small penguin but for some reason I can’t find a reference to it. But it would look so kewl at the controls of a nxt contraption.
Nxt Up, Flying Robots
Jan 13th
Hobbico NexSTAR
Chris Anderson has posted instructions for a NXT autopilot with ideas to have a flying challenge similar to the FLL (grade school) or FTC (high school) competitions that happen on the ground (mostly).
I have a lot of work already lined up to support FRC and FLL here in the Greenbelt community. But, with NASA Goddard 3 miles up the road, we also have a lot of local interest in anything that flies. And I am an ex-pilot…
Maybe this summer…
Hammer, piston, and foot power
Jan 12th
Tom Dempsey
At our first design meeting last weekend, after learning the rules for this year’s FRC competition, we (that would be Team 53*) debated the pros and cons of using a piston or a hammer as the primary ball shooter. The hammer idea had only one backer, but we’re still discussing the options. And then it occured to me, that, of course, no one knows the Tom Dempsey story. That probably includes most of the mentors. Sigh. I’m old. And wise.
* website under construction
Soccer playing robots
Jan 10th
FRC 2010 Competition Field
The FRC 2010 kickoff was yesterday with teams learning the competition rules for the season. This year the team will design and build a robot to play the game of Breakaway. There is an animated video produced by the game designers. Each team joins two others to form an alliance to play a match against another alliance. In early rounds, you have random alliance partners. Points are scored for each goal (obviously), and bonus points are awarded at the end of the match for elevating off the ground by grabbing or climbing a tower, or hanging off an alliance robot already on the tower.
In the afternoon, we headed back to ERHS and had a 3 hour design session. If it isn’t obvious, we have a LOT of issues to foresee and resolve while desiging, building, programming, and testing. All in the next 6 weeks; the robot ships to the March 11-13 regional competition in Baltimore on Feb 23rd.
Remote controlled shooterbot
Jan 3rd
I made the standard shooterbot, the second example robot in the NXT 2.0 kit, and then modified it to be remote controlled. Two motors are used to drive the bot, with the third operating the front mounted shooter. Three sensors are attached; ultrasonic mounted on the front left, color on the upper right, and the Playstation 2 controller interface on the back. The example program has the bot protect a 40 cm circle, giving both a verbal and visual warning before shooting. The first version has a few problems; it advances to give a warning but doesn’t return to the original location, and the gun is inconsistent; it sometimes jams, sometimes shoots twice. I haven’t added any remote control support yet.
Robotics
Jan 1st
Last month I volunteered to mentor students in the FIRST robotics club at Eleanor Roosevelt HS in Greenbelt. I know nothing about robotics, which they will soon discover, but I hope to learn fast and contribute some this year and more next year.
I would also like to help younger kids in grades 4-8; they have a program built around the LEGO Mindstorms NXT though I don’t know if any of the local schools have an active program. So I picked up a NXT kit as my xmas gift to the family.
Eva and I have built the first couple of easy robots, and I’ve played with a playstation remote control and some of the sensors. We’re looking forward to chasing the kittens around the house.