How to stop on ongoing motor motion using another condition?

[laurensvalk]

This was asked via email, so I've opened a discussion here so everyone can see the possible answer, and share other solutions.

A particularly valuable case is when the Lego robot encounters an obstacle and cannot complete the programmed motor movement (motion or arm movement), how to handle such a situation programmatically?

How to detect that the robot has caught something while driving and cannot continue moving?

These are typical situations, detection of which and response to them would make the robot more "intelligent".

I will be grateful for your tips.

As a concrete example, let's say we want to rotate motor A for 3600 degrees (10 rotations), but stop earlier if we see an obstacle or the motor gets stuck, so we don't keep waiting forever.

[laurensvalk]

Another way of phrasing this more systematically, is to say we want to rotate the motor

• until it reaches 3600 degrees; or
• until it gets stuck/stalled; or
• until the sensor sees an obstacle...

... whichever comes first.

There are a few solutions.

The easy way with Pybricks

You can implement this quite literally with the multitask block, and set it to complete when one task finishes:

It will do three tasks at once: rotate, wait for stall, and wait for sensor. It completes when one of them completes, which is exactly what we wanted!

Then the program continues with the rest, which is just a sound in this case.

The hard way 1 (synchronously using classic apps/solutions)

In the classic solution, you could:

• start the motor to run forever
• while not (3600 degrees turned OR stalled OR object detected):
  • wait and do nothing
• stop the motor
• use nested if/elsif/else to find out which of the three conditions caused the stop
• based on that, decide what to do next

The hard way 2 (asynchronously using classic apps/solutions)

• start the motor on one stack
• use event hats to detect objects or getting stuck
• stop the motor from the other stack
• wait for broadcast on the main stack
• depending on which broadcast was received, decide what to do.
• reset all state/mutex/lock variables so you could repeat all this again

[trwhitt]

I have a quick question that maybe you can help me with as we dabble with Pybricks. How would one do this inside of another task? We keep getting:

TypeError: 'int' object isn't iterable

Attached is a picture, thank you!

[laurensvalk]

This snippet looks fine, so I suspect the issue is coming from somewhere else. Could you share the generated code and the full error so we can see which line is causing it?

[trwhitt]

Laurensvalk!
Thanks for replying. We tried some things this weekend but still couldn't solve it yet so we kept working while avoiding multitasking. We'd really like to utilize some stall detection!

[laurensvalk]

I'm going to guess that prior_program and next_program are functions that you write yourself, perhaps in another file?

If you use multitasking, it is best to do so everywhere. Since it is being called as await prior_program, it is expecting a function of the form async def instead of just def.

[amberneb]

Thank you! We will give this a try!