Here is an early teaser image of the new 2WD Rover we are working on.
lynxmotion.com/images/jpg/thumbs/2wd01s.jpg
I plan to put an upper deck with rotate to accomodate a biped torso. This will be my take on a Johnny 5 robot kit from the movie Short Circuit.
The chassis uses our GHM-04, 7.2vdc, 50:1 motors with our new motor encoder kit
lynxmotion.com/Product.aspx?productID=448&CategoryID=11
So eventually this bot will know where it is, and (hopefully) know where it’s going. 
cool, now it would be really great if it had tracks.
Very cool! This is the best looking rover to date. Tracks would be a nice accessory upgrade in the future when it comes out.
Encoders no likey tracks, hence the wheels. Skidding like in turns with tracks is death to encoders – they get lost.
I did not know that. Well then a track upgrade would have to also include motors without encoders I guess.
I seem to remember a trailing wheel on Johnny 5; perhaps you could keep track of its rotation and orientation for use with dead reckoning.
nick
I see you have a servo on the front, with a bracket. What do you think you’ll have mounted on the bracket? SRF04? Sharp IR? other?
Hello Randy!
Sure, but that’s easy. I added this arm over the weekend. Because it has relatively short, 3.625" shoulder to elbow and 4.25" elbow to wrist, dimensions, and is load balanced, it will be quite strong. I will have a larger gripper capable of holding a soda can soon.
lynxmotion.com/images/jpg/thumbs/2wd02s.jpg
I should get the Johnny 5 robot prototyped over the long weekend.
Very nice!
I was kinda hoping you would post a pic of the humanoid torso mounted with the new waist rotate complete with long arms. 
You guys produce the best looking bot kits… period. Keep up the good work, and I will be looking foarward to new and exciting pics as your time permits.
Thanks very much for sharing with us your progress!
An appendage! Ah! I hadn’t expected that. Don’ t know why, but was expecting two arms on a torso above it.
Guess I was expecting a pan and tilt sonar, like on the first walker I got. I really liked that sonar set up. Thought that made neat “eyes” for the front of a robot, and the ability to look around… all there better.
Is there any chance that you’ll make that four wheel drive? I really like the look of that platform, it is very sleek. I think it would look even cooler with four wheels propelling it.
Any ETA on this bot? Maybe before X-Mas? 
Are you going to put something on the back of the robot?
Also, the 4WD version would be neat .
I don’t think you guys asking for 4WD are quite getting it. They already have 4WD vehicles. This one is for using with encoders. Encoders do not work with 4WD skid steer because as soon as you turn, the wheels skid and you lose position with the encoders. If you want a 4WD, buy one of the ones Lynxmotion already sells. Having built a 4WD skid steer and then wanting to get position information, I appreciate the casters in the back.
I love the caster design also. I am not a big fan of rovers, but If I were to build one, it would be the one with casters.
I see what you mean, however, the wheels don’t necessarily have to skid. The robot would simply have to drive forward, stop, drive one set of motors forward and one backwards to turn in either direction, and then continue forward or backwards. I believe encoders would still operate well with this configuration. I agree though, that it is very impractical and two wheel drive would be faster anyway.
Nope. The “one side drives forward, one side drives back” is called skid steer for a reason. The wheels skid as the thing turns. I have a 4WD bot and use this exact turning scheme and the wheels definitely skid which would lose position on the encoders. The encoders would work while you drive straight ahead, but would lose clicks while turning making it lose its position (or at least lose its orientation).
Well, actually skid steering is a little better than some have represented, and quite a bit worse than perfect.
Yes, you can go straight, stop, skid turn, and go straight again. The skid turn will often be close on the encoders to what actually happens.
However, and this is the BIG however, when you’re skidding, you leave solid traction with the ground behind. You think you turn so far, but what if you’re batteries are down, or the floor is extra sticky, etc. Just because you get an encoder count that is differential between the two sides, doesn’t mean you get exactly X degrees per encoder tick difference around the center axis of the body. Also, because of surface irregularities, you may draw off to one side or another. (Imagine one set of wheels on carpet, and the other on a well waxed floor.) You loose accuracy with every turn.
Correction using a gyro or compass can greatly improve accuracy on how many degrees are turned, but they can’t help with the non-center turning.
Anyway, Odometry with skid steering is possible, and is useful over the short haul, but it does get “off” more quickly compared to differential steering with constant ground contact.
Sorry, my bad. 
Any kind of stearing will make you turn. That is really all that matters.
