Stingray Navigation

erco

Is·the Stingray much more than a bigger, Propeller-powered BoeBot?· Agreed, it looks great and has tons of processing power (and I'd love to have·one) but since it doesn't come with encoders, it doesn't appear to be a giant leap forward in the navigation department. Open loop control of the motors means that it can't be trusted to travel in a straght line, make precision turns, or reliably get from A to B. That's really the most important thing with a robot.

I was scanning the Pololu site and came across one of their motor controllers with feedback. http://www.pololu.com/catalog/product/1392·is a single motor controller with PID feedback, adaptable to many robots, motors, etc. Seems like the Stingray would be the perfect·platform to·launch an integrated dual controller with PID feedback. Knowing the robots' specified size, weight, geometry, wheel size, motor, gearing, etc. would simplify PID calibration and make odometry much more reliable.

Clearly, users are free to roll their own encoders and implement PID, but it sure would be nice if the robots came that way out of the box.·The more common, intergrated, identical features, the easier·to build a huge Stingray community, everyone sharing apps.

BTW, this isn't a "bag on the Stingray" post, I love to see forward progress and this is a step in the right direction. I'm just curious if I'm the only one who·thinks this way, or perhaps I have missed something...?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·"If you build it, they will come."

Alsowolfman

this is a little unrelated, but here is a post where on the last page Ken Gracey describes the scribbler 2 that will have encoders and run on propeller for summer 2010.

http://forums.parallax.com/forums/default.aspx?f=25&m=388035&g=395557#m395557

I was also disappointing to see the stingray does not come standard with encoders, but i think that they will eventually add some as an accessory. I think with the propeller one would not really need a separate pid controller. you just need to devote one processor to that. I am working on an assembly program that reads quadrature encoders, calculates angle and x,y coordinates, and uses a decision matrix based on angular and linear error to determine its reaction to error, then pulses the motors on and off to maintain direction in a single processor. it is not exactly PID, but it works for my setup.

erco

Great Scribbler 2 info, Alsowolfman! Worth waiting for. I wonder what the price will be?

Several other dual motor·encoder-based motion controllers are already out there, such as:

http://hubbard.engr.scu.edu/embedded/motorcontrol/motorctrl-avr/index.html

http://www.c-stamp.com/CS110100.htm

http://www.nubotics.com/products/wc132/index.html

I'm sure that a single cog on the Propeller could do as good as any of these units. Anybody started working on this yet?

PhiPi?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·"If you build it, they will come."

Alsowolfman

oops, i guess i wasn't very clear, i am working with a propeller. am trying to write a program in assembly that can be used with only knowledge of spin. so far i have 3 functions, straight line, zero radius turn, and circle, but i need to improve my ramping, i seem to always have some wheel slipping at the end. here are some videos of it working:

here is a straight line:



and here is my circle test, it isn't just a set speed on each wheel, it is calculating how far off from the circle it is and using different speeds based on that.




my robot has excessive encoder precision at 9024 possitions per wheel revolution, so i am not sure how well my program would do with a lower precision system.

erco

WOW. Nice job on both feats! Bravo! I'm not into the Propeller yet, but I hope you're sharing your code in the Prop forum!

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·"If you build it, they will come."

Whit

Alsowolfman said...


and here is my circle test, it isn't just a set speed on each wheel, it is calculating how far off from the circle it is and using different speeds based on that.

Are you sure there is not a piece of fishing line attached to a nail·making that bot go in a circle?

That is absolutely amazing! Fantastic job.

Thanks for this thread erco. I was worth it just for this video and the Scribbler 2 info!

Regarding the encoders (from here http://forums.parallax.com/showthread.php?p=846123)...

Chris Savage said...


Yes, we were planning on wheel encoders at some point. Originally we considered the US Digital units but cost was a concern. By the time this happened there was not sufficient time to get a design finalized before the robots were ready to sell. Plus the encoders were always planned as an add-on.

I think the leap forward here is the Propeller and the mid-sized platform. Most of Parallax's robots are like the old VW Bettles. They evolve and become more and more refined over time. How we use them shapes this refinement and evolution. I think this will eventually become a great teaching platform like the Scribbler or Boe-Bot.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Whit+


"We keep moving forward, opening new doors, and doing new things, because we're curious and curiosity keeps leading us down new paths." - Walt Disney

Post Edited (Whit) : 11/17/2009 2:56:34 AM GMT

Chris Savage

Erco,

In one of the many threads discussing the release of the Stingray we mentioned that Wheel Encoders are a high priority accessory for the Stingray. There was so much demand for getting the main unit released before we had accessories for it that we decided to go ahead, knowing that clever customers would be able to adapt many types of sensors to it. But the original plan was to release some sort of IR sensor, wheel encoders, and even a line follower. Stay tuned. In the wake of all the products we have recently released we should have more time to complete the accessories for the Stingray very soon. Take care.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage

Parallax Engineering
50 72 6F 6A 65 63 74 20 53 69 74 65
·