Originally Posted by Greg McKaskle
I was at the Alamo regional helping the CSAs out. I will describe the common issues I saw, not to pick on the robots, but so that mentors and students can better prepare and recover when/if they occur.
As mentioned, it is a rough game and will uncover loose connections on a robot. I saw Anderson connectors disconnect, crimps to battery fail, ethernet cables pulled out, and digital ribbon cables come loose. Friction fits can only resist so much force. Fasteners, zip ties, and velcro are highly recommended to keep heavy components stationary and to keep connectors firmly attached.
Please trace radio power wires and check that it is wired as shown in R43 of the Robot manual. Late on Saturday, I was still seeing inventive ways of powering the radio. Most reboots I saw were not the cRIO, but the radio or entire robot. The log messages will indicate which devices rebooted during a match.
I saw more joystick issues than I expected. The DS doesn't know how many joysticks are supposed to be plugged in, so the LED on the left indicates zero or nonzero. The diagnostics screen and setup screens show individual joysticks. If your team connects a joystick during a match, they do not need to restart the DS, reboot the laptop, or reboot the robot. The procedure is to take hands off the joysticks, press F1, and resume play. If you press F1 with a joystick or gamepad off center, you are affecting the zero-calibration and may find that the robot is biased or spinning in circles. If this happens, follow the F1 procedure. Please make sure drivers are aware of this.
Please test your code using the practice mode of the DS. This sends the same commands to the robot as the field during a match. Some teams test auto and tele very well, but only in isolation. The transition between auto to tele happens once per match. It is important, so please test it.
Robot bumper rules, including color, are being enforced quite strictly. I understand that robots may be bypassed or estopped if the bumpers are sagging, broken, or molting.
I believe all of the fog-horns at Alamo were because a ball was reintroduced into the field improperly. They do not necessarily mean that the field is having technical difficulties.
Finally, if your robot stops moving on the field, there are a number of things to make note of to help identify and resolve the issue quickly.
* On the robot, observe the RSL and radio LEDs. You can often determine whether the entire robot lost power, can see the radio reboot, and can even see Anderson connectors with nothing connected to them.
* On the DS, observe the LEDs on the left side. Flip to the Diagnostics screen and observe the ping status LEDs and diagnostic messages. On the charts tab, you can see the realtime traces of CPU usage, battery level, and communications -- at least you can see the data for the time when the robot was connected.
* After the match you can review the charts data and messages using the Log File Viewer. It can be launched on the Charts tab, scroll to the latest match, typically at the bottom.
* Observe any dashboard or LCD message that the programmers may have put into the code -- a broken encoder, an impossible limit switch combo, a spinning gyro, etc.
* Remember what was taking place on the field when the issue happened. Were you pushing? Were you hit? Did the robot behave normally before the issue?
Once off the field, discuss the symptoms and observations. Determine some likely points of failure. Safely restore power to the robot, look at LEDs and cables. If a particular device lost power, tap/wiggle/tug the device and both ends of the power cable to see if you can cause it to happen again. Look for bare wire than may have shorted to the frame or another wire. Look for melted or pinched wiring insulation. Be sure to test sensors and actuators for damage. Most FRC code doesn't behave well with sensor failures. If you don't find the issue, ask for assistance.
When something happens to the robot, things can get intense. People may have lots of questions or theories. They may remind you again and again to hurry up and fix it. This doesn't only happen in FIRST. I hate to see robot failures like these happen, but watching a team effectively debug and correct issues is one of my favorite elements of FIRST. I hope this info helps.
Greg McKaskle
|