Minutes of the PARTS meeting on February 1st, 2003
President Pete Skeggs opened the meeting with a moment of silence in honor of the crew of the space shuttle Columbia who were killed in a re-entry breakup of the shuttle earlier in the morning.
Pete congratulated students who participated in, and recognized the PARTS volunteers who helped with, the ORTOP FIRST Lego League state tournament in January.
Pete spoke briefly about the upcoming PDXBOT.03. He said that our request for use of the Smith Center has been submitted, but that he still hasnt gotten a confirmation despite contacting the scheduling office at Portland State University twice in the last week. Larry Geib is coordinating PDXBOT.03 this year, and two organizational meetings have been held. Pete announced that we need a new media coordinator this year, and asked anyone interested to contact him following the meeting. Pete also asked for members who were interested in being trained as judges for the various sumo competitions. Pete suggested that ideally our judges wouldnt be competing, but they could be assigned to other events to avoid judging their own robots. Pete said its possible that well have fewer school teams at PDXBOT.03 if the school year is shortened.
Pete reported that PARTS vice-president Warren Leach had received an email from 20th Century Fox about their need for robots to be included in a movie to be filmed this summer. The set buyer interested in receiving photos and details about members unusual larger robots. If you are interested in auditioning your robot, Warren can put you in contact with him.
Pete described attending an Intel robotics symposium / marketing event for the new Intel X-Scale processor family recently. Attendees included researchers from Stanford and Carnegie-Mellon. Pete reported that Intel was getting feedback from these folks on what features needed to be included to make their robotic processors truly useful.
Pete said that one of the ORTOP volunteers works for InFocus, and informed Steve Davee about an equipment loan program. Pete will submit a request from PARTS for use of InFocus video projectors during PDXBOT.03.
Larry Geib reported that PDXBOT.03 committee members are in the process of finalizing rules for this years events. We plan to have our own copies of rules posted on the PARTS web site (as opposed to simply linking to rules on other sites) to avoid unintentional changes to rules prior to our event. He urged PARTS members who have ideas for rules to pass them along.
In the Ask Doctor Robot section of our meeting, Daryl Sandberg asked if members thought it would be possible to use a serial controller to completely replace an on-board Parallax BS2 microcontroller. Members responded that serial transmission of data would require adherence to a protocol, and that some sort of cyclic redundancy check (CRC) would be required to ensure that the data is retransmitted if it is not correctly received.
Warren Leach asked if members knew of any practical ways to modify radio control servos for 360-degree (as opposed to the more common 180- or 270-degree) rotation. The responses all dealt with the need to replace the internal feedback potentiometer, which is not designed for 360-degree rotation. One suggestion was to remove the internal travel stops and use an outboard potentiometer geared to the output shaft, along with mechanical, magnetic or optical limit switches. (This suggestion was described in the Seattle Robotics Society Encoder in October of 2000.) Other suggestions included using a 2:1 gear box to step up the output rotation (if the loss of torque is acceptable), and using only one rotation of a multi-turn sail winch designed for model sailboats.
Due to the large number of younger visitors at this months meeting, Pete asked members who had brought larger robots to show their work first (before youngsters patience ran out).
Daryl Sandberg showed his MR2 moon rover platform and camera. Since the last time he brought MR2, Daryl has added independent suspension (demonstrated when MR2 climbed over Monty Goodsons toolbox while rolling forward). Daryls chassis is made of Lexan and aluminum, powered by two 10AH 12V lead acid batteries from Wacky Willys. Daryl is using 7-pin 5A H-bridge drivers on his motors, and has worked hard to remove switching glitches. He has replaced the custom-made RTV silicone tire treads with strips of rubber doormat glued to his wheels. His old radio control transmitter gives him about a 70 foot range (enough to control the robot in his back yard while hes sitting in the living room.)
Dan Gates showed Little Johnny; he said that this robot is one that may appear in the movie mentioned earlier. Little Johnny has a HandyBoard 68HC11 controller board, and a hand grip that is actuated with a servo at the elbow, and connected to the hand through a sliding throttle cable that runs through the lower arm.
Jonathan Fant showed FRED, his robot that he hopes will eventually move autonomously around his house. Currently, FREDs IR sensor output is displayed on a PC. FREDs intelligence comes from a 10-watt processor board with 64 mbytes of RAM and a 10 gigabyte hard drive. When complete, FRED will have an arm and a domed top that will give it an R2D2-like appearance. Due to his 4-inch wheels, FRED will be limited to a flat environment, and needs additional bumper sensors and edge detectors to help him avoid obstacles that his camera might not detect. Jonathan wants to provide FRED with some human-like characteristics, and said that providing autonomous navigation capabilities is the biggest challenge.
Gene Collins showed his Mobile Bot. Like Jonathan, Gene wants a platform that can move around his home, but Genes design needs to be able to negotiate 8-1/2 inch steps. Genes platform is built of cut and soldered copper-clad fiberglass circuit board material . The resulting structure is very rigid. Gene is using DC gear motors from Jameco, and theyre mounted in easily-removable cradles. Gene got his gears years ago from Small Parts Inc.; Pete suggested Serv-O-Link Corp. as a low-cost supplier of gears, sprockets and chain. Gene demonstrated the sequence necessary to ascend a step its slow, but it works . Gene is using 32 gauge ribbon cable to pass power to his platform, but all internal connections use 18 gauge wire. When its all self-contained, the current limiting imposed by his ribbon cable (and perhaps some sluggishness) should be gone.
Mark Medonis showed one of his early projects, a head with two ultrasonic ears that is mounted on a servo . Time-difference signals from the ears are used to keep the head facing a modulated 40 kHz ultrasonic chirper. The head has a color video camera attached to a monitor. Mark says the system works fairly well, but in enclosed rooms it can be fooled by sound reflections or other noises with lots of ultrasonic content such as jingling car keys or finger snaps. Mark described another application for ultrasonic location in which sensors are mounted in each corner of a room, and locate the source of chirps anywhere within the enclosed space. Pete suggested that Marks head would be an interesting technology demonstration at PDXBOT.03.
Robert Pearson Sr. showed the LEGO robot built by the team he coached for the recent ORTOP competitions. Their robot had front line sensors, small wheels mounted on the sides for wall following and a wall touch sensor. Most impressive was the use of a rotation counter driven off a differential drive. This provided really accurate distance sensing, independent of speed or time or battery condition. Their team score 256 points in the competition.
Jonathon Spiva showed the current state of TRWNN (The Robot With No Name) . Their chassis is made of .090 aluminum, and was fabricated by Pro-Tech Industries in Vancouver, WA. The robot will have a variable-speed track drive; its driving wheels are being machined, and will be driven by 24-volt wheelchair motors . Jonathon, Robert and James have gotten some spring-loaded belt tensioners for the tracks. Control will come from a Mini-ITX motherboard running Linux.
Bruce Kroeze showed a six-legged walker robot he built based on a design in Karl Williams book Insectronics . Bruces robot uses three servos and the Mark III controller and sensor boards. Using his Zaurus PDA to send serial controls, Bruce demonstrated several walking motions.
That ended the large-robot demonstration, and the show and tell session began.
Pete gave Daryl two nicely-machined grippers (originally from Wacky Willys Surplus) for use on MR2.
Daryl showed the latest issue of Robocon Magazine, and said it was his favorite magazine in the whole world. It must be. Subscriptions are $66 a year, prepaid, through the Owajimaya store on Beaverton-Hillsdale Hwy, and you have to pick them up at the store when they arrive. Daryl announced that Tim Rohaly is taking over production of molded urethane tires.
Warren Leach reported that he had been contacted by the Opto Product Manager from Sharp Microelectronics who wondered if we use any IR sensors in our projects. Warren has previously posted info about the Sharp line of sensors to the PARTS newsgroup. Warren also had a conversation with an acquaintance who is active in the Oregon Electric Vehicle Association. He said these guys like to make conventional vehicles run on batteries, and may be good resources for PARTS members interested in the DARPA Challenge.
Larry Geib reported that we need a metal ring for the Japan-class sumo competition. A water-jet cutter is one way to cut sheet steel without heat distortion, and Larry wondered if anyone in attendance had any connections. Corey Poole suggested that the cutting might also be achieved with a sufficiently-large band saw. Steve Hampson said that even a scroll or jig saw could do the job, given enough time and the right blades.
John Roach reported that he had attended PARTS meetings years ago, and has decided (following a serious accident) to follow his dream, go back to school and learn more about robotics.
Monty Goodson and Pete Skeggs showed attendees what a micro-sumo match looks like . Montys design seemed to have more speed and better traction. It also has a unique feature: even if the robot is rocked onto its back, the wheels still contact the ring and can push forward.
Steve Davee showed his favorite micro-sumo motors from All Electronics. His new Beastie micro-sumo robot will use reject urethane tire treads (from Daryl) cemented on wheels. This robot will have more speed than his previous design. He also showed a 3D rendering from Pro Desktop, a 3D CAD program available for free download from PTC.
Dan Gates showed a Tide detergent stain remover container as a possible small robot platform. Dan is putting finishing touches on his Sumo-11 board that will fit the Mark II mini-sumo platform. This board uses a 68HC11A1 controller, and will be functional with the Interactive C language. When the expansion board is installed, this controller will have nearly all the capabilities of the HandyBoard, at a fraction of the cost. Pete recommended the PSD series of microcontroller peripherals from STMicroelectronics. Dan also announced that the web address of the Southern Oregon Robotics Club has changed to www.sorobotics.org (and www.sorobotics.com). Dan suggested that builders who are looking for serious tank-style treads check out Inuktuns MiniTrac Transporter modules.
Eric Stewart described a jewel thief circuit that he and son Eric built that powered two surface-mount LEDs for 48 hours with a dead AAA cell. Younger Eric ran across the transistor circuit on the internet.
Karl Kuchs reported that he went to the Wacky Willys location in Honolulu, HI over Christmas break. He reported that there are a lot fewer people in that store, and its still possible to walk in get the good stuff. He has abandoned a legged robot design for his robotic colony due its poor results when dead reckoning. Instead, he will use tracked robots with Tomiya gearboxes from Gordon McCombs Budget Robotics. He also has been doing reviews of computer equipment for his web site, and said that it has provided him with equipment to build a new computer, essentially for the cost of writing and posting his reviews.
Corey Poole expressed an interest in buying some high-intensity LED lighting systems from LEDTronics, but needs to combine products for other interested buyers to meet the companys minimum order dollar limit. Anyone interested should email Cory.
Tim Rohaly had some Rick Farmer PIC evaluation boards to give away. He also showed some flex sensors used in virtual reality gloves. They are basically a Kapton substrate with two layers of conductive inks printed on them. The amount of resistance at the contacts represents the amount of bend in the substrate. Tim said that the first 15 degrees of flex does not result in a linear resistance change, so many implementations have a built-in 15-degree pre-load bend. Tim is using these across the knee joints in his walking robot to sense leg obstruction.
A visiting Clackamas High School senior (sorry, I couldnt make out your name tag!) brought his robot project. The goal was to produce a robotic locker that would follow him around at school. He made a chassis out of sheet aluminum, and is using windshield wiper motors for propulsion. He decided against providing stair-climbing capability due to lack of time.
Ed Sobey and Woody have developed and present robotics education classes in the Seattle area. They are interested in ideas for kids projects and materials for a traveling exhibit.
Woody is interested in building a robotic arm to play tic-tac-toe. Anyone with similar interests can contact him.
Pete showed his revised quadrature encoder test fixture . He said that since last month he has worked to clean up the signals from the IR detectors (he resorted to the dead bug mounting method to accomplish this). He showed graphs that explain the error in signal timing and pulse position that will result from axial misalignment of the sensors with respect to the encoder disk shaft. He also passed around some generic instructions for making a toddler robot, to stimulate interest in the walking robot competition at PDXBOT.03.
Mark Medonis had some spare catalogs and battery chargers to give away.
Bob Dyer noted that a couple of the Seattle Robotics Society members were entering the DARPA Grand Challenge for an autonomous vehicle to travel from Los Angeles to Las Vegas. They are looking for participants, and plan to coordinate via email. Interested builders can check out http://www.scarabrobotics.com. Bob also had some magnetic card readers to give away.
Paul Burkey showed the SpongeBob Squad tee shirt from the ORTOP team that he mentored this year. He said that he has attended meetings of the Portland chapter of the Oregon Electric Vehicle Association for several years. He mentioned that theres a Segway and a robot at the Portland Auto Show in the Convention Center.
Karl Boe brought samples of the parts he is making on his Hardinge turret lathe . He had two forms tools made (one in high-speed steel, the other in carbide, using Electrical Discharge Machining, or EDM) so that a single infeed on the lathe cross slide would cut the entire part profile. The finish on both the brass and plastic parts was exceptional. If PARTS members need to make prototype parts and are willing to pay for tooling, Karl says hell work for pizza.
Allen McMillan showed his fashionable ORTOP Volunteer tee shirt, suitable for wear only to PARTS meetings or next years competitions.
Johnathan Fant said that he was getting noise from his GP12D IR sensors on his power supply lines, and that was causing logic problems. He has since added 10 uF and .01 uF filter caps close to the IR sensors to limit the spikes. Tim Rohaly pointed out that each sensor draws about 300 mA when it fires, so there is a lot of potential for supply bus sagging. Jonathan recommends the Electronic Rainbow 12-bit A-to-D converter that plugs into the parallel port of a PC. Ron Nucci recommended that Jonathan visit the American Association for Artificial Intelligence web site for information about robotic navigation.
The next PARTS meeting will be Saturday, March 1st, 2003.