I've had a hard time finding tracks at a resonable price as well. I've used them in a robot application before, and was not that impressed. Simple rubber tracks are too innefficient I think. One track always seems to drag more than the other resulting a need for constant compensation in software. Not that this isn't done on wheeled bots, but it was a noticably larger problem with tracks.
OAP has changed their project goals since the last time I visited the page. I respect their changes. The fact that they were attempting to (as they put it) "reinvent the wheel" really peeved me. There's some great open source robot control software on the market (Player for one). Projects like OAP need to take advantage of it. The EPIA is an excellent platform. I was consdering them, but found a mini-itx board from Global American Inc that swayed me towards the Pentium-M architecture: http://www.globalamericaninc.com/new_spec/spec2.php?id=629 That isn't the same board we purchased, but ours is similar. Global American has great prices on their equipment, and they're friendly to deal with. If you're looking for single board or industrial-type computers I suggest them, they are a pleasure to deal with. - Sebastian On Sun, 4 Dec 2005, Paul Brower wrote: > Wow, that is an impressive project you have planned. > > I am sure you have checked out the Open Automation Project. > http://oap.sourceforge.net > They seem to have an interesting project also. With a functioning > prototype. > > I have been wanting to buy a VIA EPIA motherboard and toss linux on it to > see how it performs, but just haven't had the cash. > > A few years back, www.robotcombat.com, sold a set of 'snow blower' tracks > for about $200-300. They don't seem to carry them any more and can not > find a vendor online that sells some for a reasonable price. I thought > tracks would be a good base for a linux power robot. Looks like I will > have to go wheels. > > > ~Paul > > On Sat, 3 Dec 2005, Sebastian Smith wrote: > > > Here are some of the components I'm thinking of for The Open Robotic > > Reference Platform (TORRP): > > > > The chassis for TORRP will be created out of extruded aluminum (or at > > least the prototype will be) available from 80/20 Inc. I've chosen > > extruded aluminum because of its strength, weight, and flexibility in > > configuration. I'd like to final prototype to be fully reconfigurable and > > scalable so that it can be retrofit into any research role -- think a > > robotic erector set. > > > > Motors: > > I'm currently interested in either NPC gear motors, MagMotors, or just the > > standard NPC motors. These motors powered some of the most successful > > Battlebots (for those who remember that show). I'm currently leaning > > toward magmotors because of their excellent product reviews, but they are > > a bit expensive and may be a bit more power than we need (3hp+). > > > > Motor Controller: > > I'm still thinking about this component. It will depend primarily upon > > which motors we purchase. Regardless, I'm currently leaning toward the > > RoboteQ ax2550. > > > > Battery: > > Our current robots have 3x 12V 7Amp/hr sealed lead acid batteries. They > > can keep the robot moving for about 1.5hrs. I'd like at least a 55Amp/hr > > SLA battery in the prototype. Not only should we be able to power it for > > longer, but I'd also like enough juice for more sensors. > > > > Controller: > > Pentium-M based Mini-itx. I need low-power computational horsepower! The > > pentium-m architecture will be a huge step up from the pentium-III onboard > > our current robots. With support for more sensors, and more available > > compute horsepower we'll be able to run more complex algorithms > > onboard. Since the chassis will be scalable, an onboard cluster of > > mini-itx systems is also an option. TORRPs initial design will probably > > demonstrate this capability. > > > > Sensors: > > This is a very flexible area. Initial support will include basic sonar, > > laser measurement systems (Sick LMS 200 or LMS 291), pan-tilt-zoom cameras > > (currently using Sony Evid 70, but would like to go with a usb or firewire > > option instead), and simple bump sensors. I'd like to add later support > > for GPS, LADAR, RADAR, FLIR, stereo vision systems, and microphones. This > > will depend on funding... of course. > > > > Those are most of the critical components. I'm expecting the overall > > weight of the prototype to be 75-100lbs. Not the lightest vehicle, but it > > has wheels and can drive itself! Most of the weight will be from > > batteries, which can be removed if it needs to be moved around. > > > > There are some other key features I can go into if you'd like to hear > > more. > > > > Most of the components (save the high end sensors) can be found at the > > following site: > > > > http://www.robotcombat.com/store.html > > > > - Sebastian > > > > > > On Sat, 3 Dec 2005, Paul Brower wrote: > > > > > > > > > > > On Fri, 2 Dec 2005, Sebastian Smith wrote: > > > > > > > If you're interested in the hardware I'm currently thinking of using > > > >let me know and I'll send you some links. > > > > > > I would be interested in seeing what you have planned. > > > > > > Thank you, > > > Paul > > > > > > > > _______________________________________________ RLUG mailing list [email protected] http://lists.rlug.org/mailman/listinfo/rlug
