How can I achieve a smooth, precise movement?

[Dieg0HVF]

Hi everyone, I'm competing in FLL and WRO, and I want to complete missions as quickly as possible while maintaining accuracy.

I discovered that if you use "continue movement" (which is the same as "Stop.NONE") when moving forward, and then use a turn block with a radius to turn, it does so without stopping. My problem arises when trying to move forward smoothly after that turn; the issue is that when I use "Stop.NONE" on the radius turn block, it ends up turning too far, making it imprecise.

My question is, is there a precise way to move forward, turn, and continue moving forward without having to stop?

Thanks.

[jaguilar]

Have you tried the Stop types other than NONE? Stop.NONE applies no braking so it is unsurprising that it would overshoot your target.

[dlech]

In this case, Stop.NONE only applies for an instant before the next action starts, so it should not have a physical effect. Other stop actions do have an effect, which is why Stop.NONE was added. So I think the usage here is correct.

[dlech]

What is the axel track and wheel size of the drive base?

Are you using use_gyro(True)?

It the overshoot the same every time? e.g. can you use 88 instead of 90?

[Dieg0HVF]

First of all, thank you for replying.

The wheel diameter is 62.4 mm and the tire spacing is 117 mm. We are also using use_gyro(True).

The overshoot is much greater if I set it to 90; it usually turns between 119 and 125 degrees, and it's not always the same.

After a certain distance, it corrects the angle to 90. The problem is that if I set it to turn less, then when I want to move forward, it will redirect to those degrees and not actually to 90.

And I want to clarify that the robot works normally for me; I'm just trying to find a more efficient way to move.

This is the programming, and below you can see the rotated angles.

WhatsApp.Video.2025-12-31.at.11.41.45.PM.mp4

[laurensvalk]

It looks like almost all of the weight is in the back of the robot. At full speed, it's going around the corner a bit like a pendulum with all that weight, causing the wheels to slip.

If you're saying it eventually does recover to 90 degrees it sounds like the software is working as intended, but it can't keep up with the physics :)

To help rule this out, can you try make it go very slow and see if it doesn't overshoot in that case? I'm not saying you should drive slow in the competition, but if this works, you could think about the mechanical design some more to work better at high speeds.

[Dieg0HVF]

The weight is distributed almost equally between the front and back.

When I set it very slowly, the robot still overshoots, but only by 2 or 1 degrees, and it corrects itself almost instantly.

[dlech]

The weight is distributed almost equally between the front and back.

The weight needs to be evenly distributed on each side of the axle. In other words, the wheels themselves need to be in the center of the robot. In the video, the wheels are clearly much closer to one end of the robot than they are to the center.