fix(api): Center instrument over moveable trash regardless of configuration #14413

CaseyBatten · 2024-02-01T22:41:28Z

Overview

This seeks to solve problems identified in: RSS-419, RSS-434

Rather than cover many addressable area types, we have chose to special case the moveable trash given the likelihood of an adjacent item obstruction causing an overlap in various partial tip configurations.

The solution to this is to ensure that the waypoint destination does not utilize the active channel of the pipette (such as a single column on a 96 channel) but rather the center of the destination instead.

We also override the adjacent slot checks in cases such as this, as we know that we will not overlap their slots in instances involving the trashbins because of this reconfiguraiton.

Test Plan

Ensure the a protocol switching between partial and full tip pickup drops tips within the bounds of a moveable trash the same way regardless of configuration, and does not at any point overlap with adjacent slots as identified in various collision cases.
Test Case:

from opentrons import protocol_api
from opentrons.protocol_api import COLUMN, ALL

requirements = {
    "robotType": "Flex",
    "apiLevel": "2.16",
}

def run(protocol_context: protocol_api.ProtocolContext):
    adapter1 = protocol_context.load_adapter("opentrons_flex_96_tiprack_adapter", "A2")
    adapter2 = protocol_context.load_adapter("opentrons_flex_96_tiprack_adapter", "A1")

    trashA3 = protocol_context.load_trash_bin("A3")

    tiprack1 = adapter1.load_labware("opentrons_flex_96_tiprack_200ul")
    tiprack2 = adapter2.load_labware("opentrons_flex_96_tiprack_200ul")
    partial_tiprack= protocol_context.load_labware("opentrons_flex_96_tiprack_200ul", "B1")

    pipette = protocol_context.load_instrument(
        "flex_96channel_1000", mount="left", tip_racks=[tiprack1, tiprack2, partial_tiprack]
    )

    #Perform two full tip pickup and drops, watch the overlap they drop in
    pipette.configure_nozzle_layout(style=ALL)
    pipette.pick_up_tip(tiprack1)
    pipette.drop_tip()
    pipette.pick_up_tip(tiprack2)
    pipette.drop_tip()

    #Perform two partial tip drops. Ensure they occupy the same tip drop alternation locations as the full drops.
    pipette.configure_nozzle_layout(style=COLUMN, start="A12")
    pipette.pick_up_tip(partial_tiprack.wells()[0])
    pipette.drop_tip()
    pipette.pick_up_tip(partial_tiprack.wells()[8])
    pipette.drop_tip()

This protocol should also be tested for an 8 channel pipette with two full pickups (no partial necessary at this time) to ensure that we have not caused the pipette to incorrectly center over A1 when moving to the moveable Trash.

Risk assessment

Waste chutes were omitted from special casing on this version, so as not to interrupt the current addressable areas provided by the waste chute with cover fixture. This fixture supports single channel tip drop from a 96 channel, but if we centered the partially configured 96 channel on the waste chute we would not be able to make use of the slot. Right now, the 96ch adapter is taller than the waste chute. This means a 96ch adapter in D2 could potentially obstruct the 96ch if partially offset on the Waste Chute, however we would need to verify if such a collision was even possible.

…le areas

codecov · 2024-02-01T22:46:36Z

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (af9da37) 68.14% compared to head (4257a06) 68.15%.
Report is 10 commits behind head on edge.

Additional details and impacted files

@@            Coverage Diff             @@
##             edge   #14413      +/-   ##
==========================================
+ Coverage   68.14%   68.15%   +0.01%     
==========================================
  Files        2522     2522              
  Lines       72023    72228     +205     
  Branches     9218     9314      +96     
==========================================
+ Hits        49077    49226     +149     
- Misses      20757    20787      +30     
- Partials     2189     2215      +26

Flag	Coverage Δ
app	`64.75% <ø> (+0.08%)`	⬆️
g-code-testing	`96.48% <ø> (ø)`
hardware-testing	`∅ <ø> (∅)`
protocol-designer	`37.93% <ø> (ø)`
shared-data	`75.25% <ø> (ø)`
step-generation	`86.90% <ø> (ø)`

Flags with carried forward coverage won't be shown. Click here to find out more.

Files	Coverage Δ
api/src/opentrons/hardware_control/types.py	`96.55% <ø> (ø)`
...c/opentrons/protocol_engine/clients/sync_client.py	`100.00% <ø> (ø)`
...rc/opentrons/protocol_engine/execution/movement.py	`100.00% <ø> (ø)`
api/src/opentrons/protocol_engine/state/motion.py	`100.00% <ø> (ø)`

... and 17 files with indirect coverage changes

This reverts commit 8effe0b.

SyntaxColoring

Thank you! I have one fundamental question, and a number of smaller things that follow from that.

SyntaxColoring · 2024-02-05T15:23:34Z

api/src/opentrons/protocol_engine/state/motion.py

-        destination_cp = CriticalPoint.XY_CENTER
+        if "movableTrash" in addressable_area_name:
+            destination_cp = None
+        else:
+            destination_cp = CriticalPoint.XY_CENTER


Fundamental question. From your PR description:

Rather than cover many addressable area types, we have chose to special case the moveable trash given the likelihood of an adjacent item obstruction causing an overlap in various partial tip configurations.

Can you elaborate on this?

The moveToAddressableArea command is deliberately specified to center all of its nozzles over the given addressable area—independent of whether the addressable area is a trash, or something else. We should aim for the implementation to match that specification. Are there any challenges with doing that?

In other words, I don't think there should be anything here that's conditional on the addressable area.

The current behavior is that the center of the chosen pipettes is the "critical point" that moves to the center of the addressable area. This means that when a pipette is in partial configuration, the critical point becomes the center of that selection. For example, the center of a partially configured 96 channel in COLUMN configuration for column A1 would be the middle of that column (so somewhere between D1 and E1?). That is my current understanding anyway. Whether or not we want to use the instrument center for ANY AND ALL move to addressable area commands is a difficult question to answer.

In the case of movableTrash, we specifically want to use the center of the instrument, because we're in the trash! We don't want to overlap with anything outside of the bins walls.

In the case of the wasteChute with a cover attached, we would instead want the partially configured 96 channel with single column pickup to use only its active channel as the critical point, so that it can still utilize the cover slot (unless we want to remove that support).

There may be some other examples as well, but these are the two examples related to this PR.

Whether or not we want to use the instrument center for ANY AND ALL move to addressable area commands is a difficult question to answer.

Okay, I think this is the crux of the confusion. For the purposes of what this code needs to do, we've already answered that question. The answer is "yes, use the instrument center always."

(For the broader purposes of future extensions to the deck configuration system, and generically supporting lots of fixtures and lots of pipettes, yes, it gets more difficult. But even there, I don't think the solution will be to special-case different addressable areas; it will be to make it configurable per-area, or add a parameter to moveToAddressableArea.)

In the case of the wasteChute with a cover attached, we would instead want the partially configured 96 channel with single column pickup to use only its active channel as the critical point, so that it can still utilize the cover slot (unless we want to remove that support).

As far as I know, this was never supported. We can double-check this, but for example, if a Python protocol does p96.drop_tip(waste_chute), I think it will emit a moveToAddressableArea command with "addressableAreaName": "96ChannelWasteChute", not "addressableAreaName": "8ChannelWastechute". That would be the proper use of the Protocol Engine API, as documented.

api/src/opentrons/protocol_engine/state/motion.py

api/src/opentrons/protocol_engine/state/addressable_areas.py

api/src/opentrons/protocol_api/core/engine/deck_conflict.py

TamarZanzouri · 2024-02-06T15:34:21Z

api/src/opentrons/protocol_engine/commands/move_to_addressable_area_for_drop_tip.py

+            "Whether to utilize the critical point of the tip configuraiton when moving to an addressable area."
+            " If True, this command will ignore the tip configuration and use the center of the entire instrument"
+            " of the entire instrument as the critical point for movement."


Suggested change

"Whether to utilize the critical point of the tip configuraiton when moving to an addressable area."

" If True, this command will ignore the tip configuration and use the center of the entire instrument"

" of the entire instrument as the critical point for movement."

"Whether to utilize the critical point of the tip configuraiton when moving to an addressable area."

" If True, this command will ignore the tip configuration and use the center of the entire instrument"

" as the critical point for movement."

api/src/opentrons/protocol_engine/commands/move_to_addressable_area_for_drop_tip.py

CaseyBatten · 2024-02-06T16:46:39Z

Utilizing the new instrument critical point approach we achieved the desired result of the protocol outlined in the test plan above. When dropping tips in a trash bin, regardless of configuration, the pipette correctly distributed them within the bounds of the trashbin in the expected safe tip drop locations.

CaseyBatten · 2024-02-06T21:25:40Z

Resolution to noted issue by defaulting ignoreTipConfiguraiton behavior to true. Behavior change ensures that trash bins and waste chutes always operate with center of instrument. Tested on Flex, performed as expected. Of note, covered waste chutes are already incompatible with the 96 channel, so we are safe to proceed.

Issue has been resolved.

sanni-t

LGTM! Thanks for making the changes!

CaseyBatten added 3 commits January 31, 2024 17:03

special casing of motion planning for trash and waste chute addressab…

86fb2dc

…le areas

removal of waste chute case and addition of tests

712912b

lint fixes

63d3171

CaseyBatten requested a review from a team as a code owner February 1, 2024 22:41

CaseyBatten requested review from a team and sanni-t February 2, 2024 15:35

CaseyBatten added 3 commits February 2, 2024 12:38

test case improvements

2a39a03

Merge branch 'edge' into center_partial_tip_over_trash

139cde5

hardware testing patch correction

8effe0b

CaseyBatten requested a review from a team as a code owner February 2, 2024 20:33

CaseyBatten added 2 commits February 2, 2024 15:37

Revert hardware testing patch correction

ef5d531

This reverts commit 8effe0b.

movableTrash typo correction

1e20b73

SyntaxColoring previously requested changes Feb 5, 2024

View reviewed changes

jbleon95 and others added 3 commits February 5, 2024 14:28

fixed hardware testing patch

c0181f5

Merge branch 'edge' into center_partial_tip_over_trash

6957feb

Addition of full instrument critical points

a6e9429

TamarZanzouri reviewed Feb 6, 2024

View reviewed changes

SyntaxColoring reviewed Feb 6, 2024

View reviewed changes

api/src/opentrons/protocol_engine/commands/move_to_addressable_area_for_drop_tip.py Show resolved Hide resolved

CaseyBatten added 2 commits February 6, 2024 10:59

further additions to the instrument critical point transition

87b02c9

addition of missing ignoretipconfiguration parameter

2eaffb7

CaseyBatten added 5 commits February 6, 2024 13:22

test fixes and addition of schema updates

f2e6ff9

hardware testing api patch update

8db73d8

comment cleanup

5303eee

default all tip configuration ignore for drop tip to true

784360d

schema and test updates

37e5fc8

format fix

4257a06

laviera requested review from TamarZanzouri and SyntaxColoring February 6, 2024 22:23

sanni-t approved these changes Feb 6, 2024

View reviewed changes

laviera approved these changes Feb 6, 2024

View reviewed changes

CaseyBatten merged commit b022b4b into edge Feb 6, 2024
56 checks passed

CaseyBatten deleted the center_partial_tip_over_trash branch February 6, 2024 22:27

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix(api): Center instrument over moveable trash regardless of configuration #14413

fix(api): Center instrument over moveable trash regardless of configuration #14413

CaseyBatten commented Feb 1, 2024 •

edited

Loading

codecov bot commented Feb 1, 2024 •

edited

Loading

SyntaxColoring left a comment

SyntaxColoring Feb 5, 2024 •

edited

Loading

CaseyBatten Feb 5, 2024

SyntaxColoring Feb 5, 2024

TamarZanzouri Feb 6, 2024 •

edited

Loading

TamarZanzouri Feb 6, 2024

CaseyBatten commented Feb 6, 2024

CaseyBatten commented Feb 6, 2024

sanni-t left a comment

fix(api): Center instrument over moveable trash regardless of configuration #14413

fix(api): Center instrument over moveable trash regardless of configuration #14413

Conversation

CaseyBatten commented Feb 1, 2024 • edited Loading

Overview

Test Plan

Risk assessment

codecov bot commented Feb 1, 2024 • edited Loading

Codecov Report

SyntaxColoring left a comment

Choose a reason for hiding this comment

SyntaxColoring Feb 5, 2024 • edited Loading

Choose a reason for hiding this comment

CaseyBatten Feb 5, 2024

Choose a reason for hiding this comment

SyntaxColoring Feb 5, 2024

Choose a reason for hiding this comment

TamarZanzouri Feb 6, 2024 • edited Loading

Choose a reason for hiding this comment

TamarZanzouri Feb 6, 2024

Choose a reason for hiding this comment

CaseyBatten commented Feb 6, 2024

CaseyBatten commented Feb 6, 2024

sanni-t left a comment

Choose a reason for hiding this comment

CaseyBatten commented Feb 1, 2024 •

edited

Loading

codecov bot commented Feb 1, 2024 •

edited

Loading

SyntaxColoring Feb 5, 2024 •

edited

Loading

TamarZanzouri Feb 6, 2024 •

edited

Loading