Team 399: Eagle Robotics Presents: X-1

[Jeremy Germita]

Hello All!

FIRST Team 399 is proud to present our entry into the 2012 FRC game, X-1.
X-1 represents 6 weeks of hard work, breaking barriers in terms of our traditional work flow and processes.

Unveil video (http://www.youtube.com/watch?v=1AhSDXvhZ1U)

Specs:
Drive train: 2 speed drivetrain. 15 feet per second in high gear, 4 feet per second in low gear. 8 wheel drive with 4 inch wheels. Geometry allows for traversal of any obstacle in Rebound Rumble.

Conveyor belt: Polycord conveyor belt. Ground to turret in about 2 seconds.

Turret: 340 degrees of motion. Camera controlled turret angle. Closed loop positioning for other functions and safety limits. RGB light ring on camera allows for dynamic control of camera tracking range(some colors track better at certain ranges than others)

Shooter: 6" wheels powered by two FP motors and geared for 4500 RPM. can shoot from about mid field. Closed loop velocity control using constant voltage control(not percent voltage control).

"Dropper": Our bridge/bump traversal tool. Uses mechanical linkages to transfer load into structure and not actuators. Acts as a wheelie bar in the case of forward tipping.

Length: 27.5"
Width: 37.5"
Height: 54.5"
Weight 118 lbs

http://team399.bmrd.net/aboutUs/robot/files/stacks_image_269.jpg

Thanks to our sponsors: NASA Dryden Flight Research Center, Lockheed Martin, Northrop Grumman, Boeing, JT3, ITEA, BAE Systems, JCPenney, Aerovironment, Superior Grocers, Lancaster High School, and the Antelope Valley High School district.

Also thanks to Pacific Coast Coating for giving a discount on our powdercoating and the NASA Dryden Machine shop for machining our parts!
Team 399 is both proud and honored to be a part of the East West Collaboration Project(thanks, guys!).

Team Site -- Robot Page (http://team399.bmrd.net/aboutUs/robot/index.html)

See you in San Diego, Utah, and Colorado!

[Akash Rastogi]

It has been incredible having 399 as part of the EWCP. I sincerely hope you have gotten as much out of it as you put have into the group!

The robot turned out fantastic as expected, and seeing the development of it was pure awesome. I know our team wants to especially thank you for help and collaboration with programming.

Best of luck to 399 :) You guys are a wonderful team to have worked with.

[Greg McKaskle]

Great looking robot. I'm curious if the stage spotlight messed up the tracking. The retroreflective material will return any light coming from the camera and from behind/around the camera. It looked like the spot was directly behind the robot shining on the shooting targets. Harsh. Much more difficult than anything I expect to see on a field.

Greg McKaskle

[davidfv]

Great looking robot. I'm curious if the stage spotlight messed up the tracking. The retroreflective material will return any light coming from the camera and from behind/around the camera. It looked like the spot was directly behind the robot shining on the shooting targets. Harsh. Much more difficult than anything I expect to see on a field.

Greg McKaskle

It was manual aiming and shooting at our roll-out event. Still have some programming to do this weekend.

[Mark Sheridan]

wow, that is a cool robot. Your website says you use an ultrasonic sensor for distance measurement. Could you elaborate more on its usage? Is it accurate?

[thefro526]

Glad to see that 399 has stepped their game up yet again. Hope to play with you guys at some point.

[Jacob.B]

Great robot and colors, i'm from Truck Town Thunder and we are orange and blue, as well.

[Jacob Paikoff]

Great looking robot, it looks like 399 will be a force again this year. I really, really like the dropper and I'm kinda mad I didn't think of it.

Good Luck this year.

[Dan-o]

Great looking machine. You guys have a lot to be proud of. Good luck in the competition.

[Jeremy Germita]

wow, that is a cool robot. Your website says you use an ultrasonic sensor for distance measurement. Could you elaborate more on its usage? Is it accurate?

Sorry about that. That was written a few weeks ago when we were playing around with the sensor. We have developed past the need for the ultrasonic sensors after developing some encoder stuff. Ultrasonic sensors are only really accurate if you are within 20 or so degrees of perpendicular to a wall or some other object being measured. It also gives weird readings when the sound hits a porous material such as the balls or bumpers.

Great robot and colors, i'm from Truck Town Thunder and we are orange and blue, as well.

Haha. All of our media seem to depict our purple a very dark blue.


Great looking robot, it looks like 399 will be a force again this year. I really, really like the dropper and I'm kinda mad I didn't think of it.

Good Luck this year.

Thanks. It seems it is our most versatile system. We've tested it to work as a bump/bridge traversal tool, a wheelie bar, and a mini hopper!

Great looking robot. I'm curious if the stage spotlight messed up the tracking. The retroreflective material will return any light coming from the camera and from behind/around the camera. It looked like the spot was directly behind the robot shining on the shooting targets. Harsh. Much more difficult than anything I expect to see on a field.

Greg McKaskle

From what I saw from the camera's live feed, the spotlight didn't have much of an effect over our RGB ring on board. We are testing software that dynamically changes the RGB's color and camera HSL thresholds based on target distance or target visibility

[Mr. Lim]

From your pictures, it looks like you are using Jaguars via CAN.

It also mentions you are using 2 FPs for your shooter, using speed control.

Are you using the built-in Speed PID function of the Jaguars?

If so, two questions for you:

1) How did you tune your Speed PID for the Jaguars? What are your PID constants?

2) How did you get 2 Jaguars to run in sync with each other without them interfering with each other? Did you split the encoder signal? Read the voltage from one Jaguar and send it to the second?

[Jeremy Germita]

From your pictures, it looks like you are using Jaguars via CAN.

It also mentions you are using 2 FPs for your shooter, using speed control.

Are you using the built-in Speed PID function of the Jaguars?

If so, two questions for you:

1) How did you tune your Speed PID for the Jaguars? What are your PID constants?

2) How did you get 2 Jaguars to run in sync with each other without them interfering with each other? Did you split the encoder signal? Read the voltage from one Jaguar and send it to the second?

Thanks for noticing. Before I explain anything, I'd like to thank you for all the work and information you've posted on CAN Jags and Velocity control.

Following the advice from one of your posts, I decided to offload V PID to the cRIO. However, we are still using the encoder port of the Jag(one encoder to one jag). Unfortunately, we found major inconsistencies in the jag's velocity data(+/- 20%). As a slight workaround, we have our software calculate the derivative of jaguar.getPosition rather than letting the jag do it.

Since all of the processing is on the cRIO, it was pretty simple to sync the two motors. I simply took the output from the calculation and set it to the jaguars while using SyncGroups.

Thank you, and if you have any more questions, ask away!

[Mr. Lim]

Thanks for noticing. Before I explain anything, I'd like to thank you for all the work and information you've posted on CAN Jags and Velocity control.

Following the advice from one of your posts, I decided to offload V PID to the cRIO. However, we are still using the encoder port of the Jag(one encoder to one jag). Unfortunately, we found major inconsistencies in the jag's velocity data(+/- 20%). As a slight workaround, we have our software calculate the derivative of jaguar.getPosition rather than letting the jag do it.

Since all of the processing is on the cRIO, it was pretty simple to sync the two motors. I simply took the output from the calculation and set it to the jaguars while using SyncGroups.

Thank you, and if you have any more questions, ask away!

Thanks for sharing the info Jeremy! I was hoping there would be a great "CAN Jaguar Velocity Control w/ 2 Motors Success Story" here. However, you wouldn't be the first team to ditch the on-board Jag PID for a more flexible software approach. I don't think you'll be the last either...

Best wishes and good luck in 2012! Hopefully we'll cross paths in St. Louis!

[delsaner]

The bridge manipulator, the elevator, the rotating shooter; all of the components seem absolutely flawless. Good luck at the competitions!!

[EagleEngineer]

Man that thing is amazing, i cant wait to compete against you guys in San Diego. :)

[Ekcrbe]

wow, that is a cool robot. Your website says you use an ultrasonic sensor for distance measurement. Could you elaborate more on its usage? Is it accurate?

We also tried an ultrasonic sensor, but it only reached about 14 feet. I don't know if it was the same one, though.

[nlknauss]

Beautiful machine making beautiful shots. The work you've done on the camera has really paid off. It seems like this will be a year where those who've developed a targeting system will really shine.

Congrats,
Nate

[Tylernol]

Looks like a great machine you guys have built. See you at the Utah Regional!

[reymac]

This looks good. Very well integrated. I also like the use of color. I too was having some trouble of figuring out if the color was blue or purple.

[Wizzard of Izz]

I can't wait to see it at the Utah Regional. Good luck this year.

[Andrew Lawrence]

I noticed in your description you said that different colored LEDs worked differently at different distances. Could you elaborate on that? Which ones worked best closer, farther, from an angle, etc?

Also, great job at the San Diego regional! You were unfortunately plagued by problems, so hopefully we'll see 399 pull a victory from Utah.

Thanks.

[Jeremy Germita]

I noticed in your description you said that different colored LEDs worked differently at different distances. Could you elaborate on that? Which ones worked best closer, farther, from an angle, etc?

Also, great job at the San Diego regional! You were unfortunately plagued by problems, so hopefully we'll see 399 pull a victory from Utah.

Thanks.

Thanks for the kind words.

From our testing, we've found these results:
Close range(< 5 feet): Red. Red isn't as intense as the rest of the colors so it doesn't wash out at closer ranges. However, it gets very dim very quickly as range increases.
Mid-Near range(4'<distance<16'): Green. Green works great at most useful ranges. However, at the close range, green washes out easily. At the far range, green starts to get a bit dimmer. Depending on your vision algorithm, this may or may not affect it.
Mid-Far Range(5'<distance< ~36'): Blue. Blue works really well at most ranges as it appears to remain a consistent shade of blue(to the camera) at most ranges. The only thing is that blue washes out at short ranges.
*"wash out" == appears to turn whiter

If I were to choose one color to use, I would pick green or blue. But I'd lean towards green as it fits our robot's colors better.

Another possible purpose for the ring light pre-match would be as a kind of a signal. Like have it default to white when there are no comms and switching to your preferred color upon gaining comms.

Thanks,
Jeremy

[Andrew Lawrence]

Thanks for the kind words.

From our testing, we've found these results:
Close range(< 5 feet): Red. Red isn't as intense as the rest of the colors so it doesn't wash out at closer ranges. However, it gets very dim very quickly as range increases.
Mid-Near range(4'<distance<16'): Green. Green works great at most useful ranges. However, at the close range, green washes out easily. At the far range, green starts to get a bit dimmer. Depending on your vision algorithm, this may or may not affect it.
Mid-Far Range(5'<distance< ~36'): Blue. Blue works really well at most ranges as it appears to remain a consistent shade of blue(to the camera) at most ranges. The only thing is that blue washes out at short ranges.
*"wash out" == appears to turn whiter

If I were to choose one color to use, I would pick green or blue. But I'd lean towards green as it fits our robot's colors better.

Another possible purpose for the ring light pre-match would be as a kind of a signal. Like have it default to white when there are no comms and switching to your preferred color upon gaining comms.

Thanks,
Jeremy

Thanks a bunch! I appreciate the information!