I think almost everyone has had an autonomous where it just kinda did something you can’t explain. We’ve had issues where if we didn’t kill the switch on the robot on autonomous it would have cost alottt of money. I remember seeing a thread about we should have an instant kill to stop a robot during autonomous if something bad happens? I can’t find it
It sounds like a good feature to have. It’s horrible to watch a robot go berserk in autonomous and completely self-destruct and/or destroy other robots on the field. Recently I was at a tournament where a a kid insisted he knew what he was doing with the code and ended up crashing his poor excuse for a robot into some well-built high school bots. He was given a warning, but then instead of disabling his wretched code, he did it again on the next match. Later he went to pieces when the refs told him to cut it out. Anyone who saw him putting his robot on their field for a match simply sank into despair.
A kill switch would have been nice to have that day. And one for the robot, too.
Glad I am not the only one who thinks so. https://www.youtube.com/watch?v=y7rzIwrEqpw
To the OP remember that interaction with the robot is illegal. This has been clarified to even include disconnecting your joystick or turning it off in an effort to abort an autonomous routine. For practice I would suggest having a fast way to kill the robot and make sure you only put an autonomous on the competition field if you are willing to face the the consequences of the routine you created. The answer is obviously TEST TEST TEST TEST.
Another thing to consider is the use of watchdog timers. For example, if you know it should take no more than 1 second for your arm to rise up, then if the arm has not achieved that height after 1.2 seconds, have the code shut itself down rather than risk the robot twisting itself and continuing on to do even more damage, etc. It can also help prevent the motors from tripping out their PTCs in the event something simply gets hung up and the motors stall. Of course, to effectively use watchdog timers, you need sensors of some kind, like pots.
Our autonomous is great, but we can test 100 times and there can always be that ONE time. We’ve test ours once 6 times in a row and it was perfect and then the next it was like 6 foot off and going into the wall, and we tested again and it was perfect. That’s the type of stuff I’m talking about.
Our autonomous won’t move on the next step until a sensor shows that the peg is in there, if a button is not pressed when the peg is in the intake, the robot will not continue on until the button is pressed. So we really don’t have to worry but we’ve also had times where it has ran two of the same autonomii at the same time.
I’m not directly talking about our robot, I mean in general like FullMetalMentor said, hurting other robots or itself it would be cool to have a button like on the tether station that you press and it kills that one robot for the 15 seconds.