PSA: Secure your battery

[efoote868]

This may be the last thing from your mind right now, but it still haunts my memory. Make sure your robot battery is secure, and the design is fool-proof.

Feel free to share your team's stories of unsecured battery mishaps.

TechHOUNDS Story time:
https://youtu.be/4uIx69RdvsY
2006 Newton QF4-1. The eventual Newton Champions and Einstein runner-ups play against the lowly 6th seed. This match is over after autonomous, the 3rd seed outscoring the opposing alliance 58-3, winning 119-69. Hope isn't lost, because the 6th seed knows they were able to outscore the opponent 66-61 in teleop. They know that all they need to do is play defense during autonomous...

https://youtu.be/Kw-ryE_8nas
4 seconds into autonomous, 868 smacks into 968, followed by 66 hitting 25. The red alliance only scores 2 balls, starting with a much more manageable 16 point advantage.
Tele-op period, and whats this? 868, the TechHOUNDS aren't moving! 10 seconds of elation now turns into season ending heartbreak. It was later determined that the battery was improperly installed, came loose and knocked open the pneumatic air dump, causing the robot to sit and never move the rest of the match. The very feature for a strategy that won our team the Xerox Creativity Award on Einstein became our Achilles heel with a loose battery...



Please don't learn the hard way that an unsecured, loose battery will lose matches.

[GeeTwo]

Thread from the Spring.

Originally Posted by Nate Laverdure

Secure your battery well, or it will fly like a fish.

We've fortunately found these sorts of problems in practice, especially last year crossing defenses. This year, the two hardest cases will be collisions between robots and falling off the rope. There were several times last year I saw robots leave batteries on the field, and shut down when the Andersons pulled apart. Plus some more on video.

While not the same level of safety hazard, also ensure that your battery connections at the terminals, through the Andersons, main breaker contacts, to the PDP, and all the crimps are tight. A flicker there will leave you sitting still for most of a match.

[Kevin Sevcik]

57 managed to lose Great Lakes in 2007 thanks to a bad battery strap. 2007 was the last year bumpers were optional. We had a cramped design so we could fit two beautiful wide ramps on the sides of the robot, so we actually had to remove a bumper to change batteries. Inspectors still wanted a strap, so a mentor went out and bought one. A 1" wide, somewhat elastic velcro strap, instead of the usual 2" wide stiff ones I was used to us running. Thursday and Friday were bad for us, because we blew up banebots transmissions driving the ramps. Had that all fixed Saturday, ended up the darkhorse 2nd pick for #1, 67 and 1114. We destroyed #8 in quarters.

Then this happened: 2007 Great Lakes Semi 1-1

That's us trying to move with a dead battery I was told was charged. Took forever to troubleshoot that, since I skipped that diagnostic, being told it was charged. (This was pre FMS having good battery data, too.) Not wanting to burn a time out, I told the students to slap in a new battery, and not bother with the bumper.

Which lead to this:2007 Great Lakes Semi 1-2

That's the battery slipping past the velcro strap and flying out the back of our robot. Then eventually getting unplugged, because that wasn't an instant disable back then.

So yeah. Batteries are heavy. Don't expect something like a veclro wrap meant for as a cable tie to hold them in. Or actual cable ties. Bolted down aluminum or beefy 2" wide non-elastic velcro is the way to go.

And then there was that time in 1999 when our arm caught the battery cable that was sticking up and pulled it out as soon as we started the match...

[mathking]

In 2006 we had a match where our very good battery box had been taken off (to access a ball that had gotten stuck in the robot) and when it was put back on the bolts were just threaded on and not tightened. Which meant that during the match the whole battery box was sliding back and forth about an inch. With about 30 seconds to go it decided "now" and ripped free from its mount.

I will also lay out the caution to make sure that your connections are not in danger of being ripped off during a match. In 2014 at Crossroads we were pursuing a ball (with our grabber extended) when an opposing robot made a deft move in front of our robot, cutting off our robot and drawing a foul (deserved) on us. Unfortunately they had mounted their battery exposed on the outside edge of the robot, above the bumper with the battery cables running straight up from the battery. Our point of contact was the exposed battery cables, which were crushed. We did get a foul but they were powerless for the last minute plus of the match. Afterwards our drive team apologized and shortly after that they came to our pit to give us the destroyed cable "in thanks for pointing out the design flaw in their robot." I believe they then protected it with lexan.

[Nessie]

I have seen robots disconnecting their connections while driving during auto/tele.
I have seen robots dragging their batteries behind them before getting emergency shutdown by FTA.
I have even seen robots destroy themselves from the inside due to a loose battery bouncing around during a match.

I witness a battery mishap at least once per regional(Usually more). Rookie teams are often the offenders but every once and a while an experienced team surprises me.


Moral of the story, secure the living everything out of your battery. Even just for a practice bot/prototype. It's significantly less time consuming and expensive to engineer a strong battery mount then to fix a busted PDB/DSC.

*Pro-tip, a zip-tie around the Andersons connection is a great failsafe. Check your alliance partners battery mounts/connections before a match and suggest the zip-tie if you have any doubts. Additionally, the extra set of eyes on each other's robots never hurts. (Yes even during qual's)

[MetaEngineer]

246 launched a battery out of the bot at least twice during drivetrain testing in 2016 before the electronics layout was finalized. Straps of duct tape were acceptable for sedate drive code tuning but not when a high speed robot did a very rapid spin in place.

The poor practice battery got ejected at high speed when the tape came loose, skidding a fair distance along the floor. We used even more tape to strap it back down again, so the second time it just ripped the duct tape in half as it went flying.

We had the coders dial back the maximum turn rate, as in addition to launching batteries it was too fast and twitchy for maneuvering. The final design had a close fitting cradle of angle pieces with velcro strap. A pain to install and remove the batteries, but we never had a loose battery in a match.

[bobbysq]

I should also point out that it's easy to keep a battery mounted. All you need is to buy or 3d print one of these and then bolt down some velcro straps on the sides of your battery. You can also have a tight fitting enclosure to put it in and you don't even need the plastic bit. (Not too tight though, our 2015 bot is responsible for some nasty scrapes on a few of our old batteries )

[flemdogmillion]

Originally Posted by MetaEngineer

We had the coders dial back the maximum turn rate, as in addition to launching batteries it was too fast and twitchy for maneuvering.

Lowering the turn rate down from 1 a little is usually a good idea, but it's also good to turn on SquaredInputs... and strongly secure your battery, because using SquaredInputs delinearizes the inputs in a way that allows better control at low speeds, while still keeping the maximum speed at 1.

[Connor McBride]

The way we mounted our battery for our 2014 robot was just horrendous. when people say that they put blood sweat and tears into their robots they mean it figuratively. We mean it literally. There was blood. There was sweat. There were tears.

[MetaEngineer]

Originally Posted by flemdogmillion

Lowering the turn rate down from 1 a little is usually a good idea, but it's also good to turn on SquaredInputs... and strongly secure your battery, because using SquaredInputs delinearizes the inputs in a way that allows better control at low speeds, while still keeping the maximum speed at 1.

This was with scaled inputs already I believe. An 18+ fps top speed might have had something to do with it. Once we were done tuning the PIDs it was quite easy to drive given how quick it was.

[jvriezen]

Some tips from my experience as a robot inspector for years and years...

1. When using velcro as part of your method to secure your battery, use a continuous loop of velcro that passes through two holes in your belly pan or other structural member. Relying on an attachment of velcro to your structure with adhesive or a screw, nut/bolt, rivet etc. usually results in a weak spot in the system, weaker than the velcro hook/loop interface.

2. Also with velcro, make the overlap of the straps as long as possible to maximize the area where the hook/loop connnection is made. Wider straps are also better, but a 3/4" strap is usually sufficient if the overlap is across the longer dimension of the battery (i.e. not just across the top of the battery.)

3. After securing your battery, test it by grabbing the battery and vigorously pulling, twisting, shaking it to see if it moves or comes out. If it comes out or has enough 'slop' so that it can bang against hard surfaces, it is not secure enough.

4. Finally, I suspect a lot of batteries that come loose on the field are due to improperly securing the battery even though there is a good securing system. Not tightening down straps or restraints correctly can make the best system useless.

5. Bungy cord based systems are often not adequate. They can be, but often is is far to easy to stretch the bungy with a battery that experiences sudden deceleration due to a hit or fall.

6. Your battery must be secure from movement in all directions-- a deep battery box with 6" high sides with an upright battery that can be lifted out is not secure, even if you aren't planning on climbing-- even if it is a flat floor this year.

7. At events I volunteer at, there is almost always an inspector in the queuing area checking the batteries on every bot to assure they are secure. Its amazing how many are not secure (at least to my standards and my LRI's standards) despite having passed inspection. Other times, teams just forget to close the restraint.

[MrRoboSteve]

If you're looking for a suggestion on a strap for your battery, these are widely available and work well when fastened in place.

I hate seeing robots dragging their battery around.