This form will be used to find out the most common problems are and solutions that each team found. Please answer to the best of your ability. Problems may include mechanical, electrical, programming or anything else.
We get to do the survey over and over for every time it quit working? 
1 Like
What if it happened midgame and endgame? What would you select as it is a ā1 answerā question and not a āselect all that applyā
Bringing this post back to the top because Iām interested too. As to comments aboveā¦
1- there can be pleasure in failure
2- probably the first time it occurred, because thereās a spot where you can select āyes, happened againā
2 Likes
Weāre pretty good at building robots that find new, creative ways of stopping each time.
5 Likes
Weāve documented our failures since 2013 if you care to see that information here. The information for 2017-2019 is much better than the preceding years.
2 Likes
Here is a presentation I made about this topic a few years ago, with some additional discussion: Five Fatal Mistakes: Why your robot is dead on the field
Iām interested in seeing how the survey results compare.
1 Like
If you ask just about any CSA, they can tell you that the most common cause by far for robots dying on the field is loose electrical connections. All of the wires between the battery, SB50 connector, main breaker, PDP, VRM, roboRIO, and radio need to be rock solid. If any of them disconnects or shorts even for a few milliseconds that can cause your roboRIO or radio to restart, which take ~30 seconds and ~90 seconds, respectively. @forbesās presentation has some good tips on how to ensure bulletproof connections. If you want to prevent problems at competition though, you really should be manually checking the wiring before and after every match to make sure nothing has come loose.
5 Likes
Following up on this, Iām not sure how much weight Iād put into your survey. Having worked with countless teams, some are fully engaged and wanting to understand the problem. Others are wanting a replay for the field fault and wonāt listen to reason. Your survey will be littered with both.
Electrical connections, above all else, are the root cause for robots dying. It can be a wire thatās gone bad and a little bump kills continuity. It could be a loose connection. It could be loose fuses. It could be an arm that swings down (yours or another robot) and hits your breaker.
But, Iād almost bet on an electrical issue right from the start. Then, Iād let the logs/robot tell me otherwise.
3 Likes
its usually because robot.exe has stopped operating
#bring back windows XP
3946ās STEAMworks robot stopped several times because the programmers put a while() loop in the command to hang a gear. But yes, Iām with @JeffB. Start off assuming itās electrical (but be ready to be wrong) and youāll find the problem faster on average.
As a tip for finding problems like these, watch the RSL when the robot dies.
If it continues to flash, electrical is less likely to be the cause. Look for oddities in software.
If it goes solid for a while, you lost power momentarily. Look for shortages
If it goes off and stays off, check that breaker and then any/all connections if it wasnāt flipped.
1 Like
If the RSL goes solid and is lit, doesnt that meant the robot has power, is disabled or lost comms but battery is still connected?
Yes. But this is also the state after a momentary loss of power (because after a loss of power the robot is disabled, it lost comms, and the battery is connected )
3 Likes
My interpretation was the robot lost power completely, as in the battery disconnected. Thanks
Had a team at an offseason event this last fall that kept dying on the field. It took them a minute to get back up and running each time. I was thinking radio rebootā¦ then when I got a close look when it happened to then again, I realized that the whole robot was shutting off for a partial second. Had the team start at their main breaker with a wire connection test.
Before their next match, theyād found and fixed a loose wire in their main power wiring.
Why do robots stop moving?
Sometimes, itās software. (Reference: Einstein 2012 Report for a particularly visible example.)
Sometimes, itās a mechanical issue. Usually a thrown/broken chain or belt.
Sometimes, itās driver error. Triggering climbs early was a relatively common robot-stopper last year, for example.
Sometimes, itās setup/design error. See previous line.
But for probably 80% of the total issues, itās something electrical. For most of those, itās going to be a niggling loose connection in the radio or RIO power. Or maybe the battery just got tossed out of the robot (and if Iām inspecting your robot, I will probably try to pick up the robot by the battery so this easily-preventable failure can be prevented!).
1 Like
I was serving as an RI at one of the last Regionals in Texas in 2018. After getting all the teams through inspection, the RIās were asked to help teams troubleshoot their robots. In the space of 45 minutes, I intermittent shorts between the positions of the Weidmuller connectors on the VRMs and PCMs on three different robots. The intermittent shorts were caused stray strands of the wires that did not make their way into the Weidmuller connector and stuck out sideways. In all cases, the teams had stripped too much insulation off their wires leaving a section of exposed conductor that was perfect for the stray strands to make contact with. Note that the way the Weidmuller connectors are used, the positions alternate between positive (or output) and ground. Thus a short between any two adjacent positions will cause a problem.
If the RSL is solid during a match instead of flashing, the most common reason is it lost power momentarily. (This also needs to losing comms as losing power resets the radio). While the radio and/or roboRIO are coming back up, the light is solid. Once theyāve rebooted, the light starts flashing again.
Itās a bit generalized to say it lost power completely. It can be a complete loss or enough voltage dropped to reset the devices (brownout). Itās important to note a complete loss can happen fast enough the logs will make it appear as if itās low voltage (if it canāt report voltage before dying, the last report is a lower number)
If itās a complete drop, itās almost certainly an electrical connection. If itās a partial drop, it might be the instance where the logs didnāt catch the āzeroā reading and it could be related to an increased current draw. The latter is generally found by watching the match and finding what was happening along with seeing a voltage drop.
1 Like
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.