bug(api): Two pipettes in use, dispensing in same well; second pipette dispensing moves diagonally downwards.

[Matt-Zwimpfer]

Overview

When two pipettes are used in a protocol (not simultaneously, but one after the other), each with their own aspirate and dispense commands, but with a shared destination well/tube, the last pipette to dispense descends diagonally into the destination well/tube.

Steps to reproduce

Create a code with the separate aspirate and dispense functions using two pipettes that dispense into the same well one after the other. Example code:

Current behavior

Aspiration steps are completed correctly and the dispense by the first pipette into the shared destination well/tube is successful, but during the dispense step of the second pipette, the second pipette moves diagonally downward towards the destination well/tube. Video of example protocol running: https://www.youtube.com/watch?v=i6QPtzjRDmM&feature=youtu.be

Expected behavior

The second pipette should move in the XY plane first to position itself above the top-center of the destination well/tube, without any movement in the Z-direction, and then descend in the Z direction to dispense into the destination well.

[mcous]

Shoot, theorized this bug a few weeks back from reading the code but never got around to testing it. Glad to have the repro.

Using a pipette sets the ProtocolContext’s location cache, but changing the active pipette does not clear the cache, so the system thinks pipette number 2 is already inside the target well and it issues a direct movement rather than an arc.

Issue has already been accounted for and avoided in the newer ProtocolEngine, but should probably be fixed in the currently thick protocol API, too.

[beniroquai]

Hey, is there any update on this? I face the same issue when performing the following:

    for i_col in range(i_experiment, N_cols):
        for i_steps in range(2):
            pipette.dispense()  
            #print("Remove liquid "+experiment_methods[i_row_bluedye]+" - column: "+str(i_col))

            '''
            Go to the microscope and dispense 
            '''    
            #get position of the first well of the microscope
            center_location = plate.columns()[i_row][i_col].center()
            
            # offset-coordinates (only calibrate it once)
            # dx=-6; dy=2; dz=150

            # go to well
            adjusted_location = center_location.move(types.Point(x=-6, y=2, z=135))
            print(adjusted_location.point)
            pipette.move_to((adjusted_location))
            time.sleep(.51)
            
            # go to down and aspirate 
            adjusted_location = center_location.move(types.Point(x=-6, y=2, z=88))
            print(adjusted_location.point)
            pipette.move_to((adjusted_location))
            pipette.aspirate(V_aspirate_blue,adjusted_location)
            time.sleep(0.51)
            
            # move up
            adjusted_location = center_location.move(types.Point(x=-6, y=2, z=135))
            print(adjusted_location.point)
            pipette.move_to((adjusted_location))
            time.sleep(.51)

            '''
            Go to the liquid and aspirate some stuff
            '''
            #get position of the first well of the microscope
            pos_reagent = plate2.columns(1)[0][1].center()
            dx=0; dy=-3; dz=-15 # offset-coordinates (only calibrate it once)

            # Get a location 1 mm right, 1 mm back, and 1 mm up from the center of well A1.
            adjusted_location = pos_reagent.move(types.Point(x=dx, y=dy, z=dz))
            print(adjusted_location.point)
            pipette.move_to((adjusted_location))
            pipette.dispense()
            pipette.blow_out()

            # move head up
            # Get a location 1 mm right, 1 mm back, and 1 mm up from the center of well A1.
            adjusted_location = pos_reagent.move(types.Point(x=dx, y=dy, z=130))
            print(adjusted_location.point)
            pipette.move_to((adjusted_location))

where the aspirate causes this diagonal motion. A second pipette is mounted too but not yet moved.

I get this issue:

I think it's complaining about reaching the lower boundary - even though it should be not even reaching it, or?

protocol is initialized like this:

# Connect to the opentrons robot
protocol = opentrons.execute.get_protocol_api('2.9')
protocol.home()

# setup the pipets
pipette = protocol.load_instrument('p300_single', 'right')

The behaviour is gone if I remove the aspirate().

Robot information:

Robot Name:
 
Robot

Firmware Version:

v1.1.0-25e5cea

Server Version:

4.2.0

Supported Protocol API Versions:

min: 2.0, max: 2.9

Thanks!

[beniroquai]

@mcous

interstingly, this diagonal motion is not hapenning in the second loop here:

protocol.home()
i_pipette = pickup_fresh_pipette_tip(i_pipette, pipette, tiprack)
for i_steps in range(2):
            pipette.dispense()  
            #print("Remove liquid "+experiment_methods[i_row_bluedye]+" - column: "+str(i_col))

            '''
            Go to the microscope and dispense 
            '''    
            #get position of the first well of the microscope
            center_location = plate.columns()[i_row][i_col].center()

            # go to well
            adjusted_location = plate.columns()[i_row][i_col].center().move(types.Point(x=-6, y=2, z=135))
            print(adjusted_location.point)
            pipette.move_to((adjusted_location))
            
            # go to down and aspirate 
            adjusted_location = plate.columns()[i_row][i_col].center().move(types.Point(x=-6, y=2, z=88))
            print(adjusted_location.point)
            #pipette.move_to((adjusted_location))
            pipette.aspirate(50,adjusted_location)
            
            # move up
            adjusted_location = plate.columns()[i_row][i_col].center().move(types.Point(x=-6, y=2, z=135))
            print(adjusted_location.point)
            pipette.move_to((adjusted_location))

            '''
            Go to the liquid and aspirate some stuff
            '''
            #get position of the first well of the microscope
            pos_reagent = plate2.columns(1)[0][1].center()
            dx=0; dy=-3; dz=-15 # offset-coordinates (only calibrate it once)

            # Get a location 1 mm right, 1 mm back, and 1 mm up from the center of well A1.
            adjusted_location = pos_reagent.move(types.Point(x=dx, y=dy, z=dz))
            print(adjusted_location.point)
            pipette.move_to((adjusted_location))
            pipette.dispense()
            pipette.blow_out()

            # move head up
            # Get a location 1 mm right, 1 mm back, and 1 mm up from the center of well A1.
            adjusted_location = pos_reagent.move(types.Point(x=dx, y=dy, z=130))
            print(adjusted_location.point)
            pipette.move_to((adjusted_location))

Is there a quick fix for it?

protocol.home() is in a different cell inside a juypter notebook

UPDATE:
The diagonal motion randomly happens also in the first iteration of the loop...