Chain drivebase movements together without stopping

[FLL-Team-24277]

Is it possible to chain movements together so that we don't have to decelerate and accelerate between movements? For example, could I run a drivebase.straight() followed by a drivebase.curve(), where if the speeds were the same for both movements, the robot would not have to stop between them. I am using a spike prime if that matters.

[dlech]

This is what then=Stop.NONE does.

https://docs.pybricks.com/en/latest/parameters/stop.html#pybricks.parameters.Stop.NONE

[FLL-Team-24277]

I was hoping it was something that simple! Thanks :)

[FLL-Team-24277]

I got to do some testing with Stop.NONE, and it makes our robot runs a lot faster. Very cool! Buuuuuuttttt, there's a problem. Sometimes, and I can't figure out when it happens, the turns go crazy. We are using a gyroDriveBase with the gyro on (we haven't upgraded to the new version just yet). What we did was chain a Straight, Curve, and Straight together, followed by a Turn. And what we saw was a couple of things, Sometimes the Curve would go wildly with a tighter radius than it should. Or following the curve sometimes the Straight would go in some completely random direction. I tried printing out the imu headings at some key places, and they frequently wouldn't make sense. Like I could print the heading at the start of the second Straight, and again at the end of that Straight, and get very different readings. I also tried drivebase.reset at a few diferent places in the code and that did not seem to help.

So, short of not using the Gyro, which I have not tested, is there anything I can do about this?

[dlech]

we haven't upgraded to the new version just yet

I would suggest trying it. There are lots of fixes between V3.2 and V3.3.

[laurensvalk]

Sometimes the Curve would go wildly with a tighter radius than it should. Or following the curve sometimes the Straight would go in some completely random direction.

No matter the software, physics will still win. If you curve tightly and then go straight, it may still be turning.

is there anything I can do about this?

If you can, try to reproduce the odd behavior on a simple drivebase like the default Spike drivebase.

If you have a video and some code, we can have a look 🙂

[FLL-Team-24277]

I'm going to have to wait until next week to upgrade to 3.3. We have our state tournament this Saturday and the team would kill me if I made a change like that right now :)

I do have video of the robot misbehaving, but I don't have any of it running correctly, which you will need for comparison. I'll see if I can get some good video this afternoon.

For now, I think we will still Stop.BRAKE after every move, and save the fancy Stop.NONE for the next tournament.

[laurensvalk]

For now, I think we will still Stop.BRAKE after every move, and save the fancy Stop.NONE for the next tournament.

Sounds like a plan 😄

[afarago]

@FLL-Team-24277 would be happy to collab and share experience with you.
We have tried it, however due to the distance_control / heading_control limits & target_tolerances not finishing the movement resulted in highly less reliable movements --> therefore we only used it in places where there is a clear hardware alignment to the model closeby or with an "S-turn" maneuver.