feat(avoidance): improve avoidance target filter #2329

satoshi-ota · 2022-11-19T05:00:22Z

Description

add new filtering process in fillAvoindanceTargetObjects()

Judge whether the object is parking vehicle on road shoulder

In the original implementation, if the length between the object's center and the center line of Ego's lane is less than threshold_distance_object_is_on_center, avoidance module determined that the vehicle was not a roadside parked vehicle and don't excute avoidance.

In this PR, not only the length from the centerline, but also the length from the road shoulder is calculated and used in the filtering process. It calculates the ratio of the actual length between the the object's center and the center line shift_length and the maximum length the object can shift shiftable_length.

$$ l_D = l_a - \frac{width}{2} $$

$$ ratio = {l_d \over l_D}$$

$l_d$ : actual shift length
$l_D$ : shiftable length
$l_a$ : distance between centerline and most left boundary.
$width$ : object width

# default
object_check_shiftable_ratio: 0.6 # [-]

The closer the object is to the shoulder, the larger the value of $ratio$ (theoretical max value is 1.0), and it compares the value and object_check_shiftable_ratio to determine whether the object is a road-parked vehicle.

If there is no road_shoulder lanelet in the map, sum object_check_min_road_shoulder_width to $l_a$, assuming that the shoulder exists.

object_check_min_road_shoulder_width: 0.5 # [m]

In addition, it is assumed in the filtering process that there are no parked vehicles within the intersection. Therefor, the filtering flow is as shown in the following figure.

Tests performed

green cube: NOT AVOIDANCE TARGET
red cube: AVOIDANCE TARGET

# default
object_check_shiftable_ratio: 0.6 # [-]

Since the object's shiftable ratio is 0.49 (< 0.6), it is determined that the object is not a parking vehicle.

# default
object_check_shiftable_ratio: 0.0 # [-]

If the value of object_check_shiftable_ratio is set to 0.0, all vehicles are judged as road-parked vehicles.

The module determines that any vehicle in the intersection is not a parked vehicle, regardless of the value of shift_length and shiftable_ratio.

Notes for reviewers

have to checkout tier4_autoware_msgs to this branch. (tier4/tier4_autoware_msgs#64)

Pre-review checklist for the PR author

The PR author must check the checkboxes below when creating the PR.

I've confirmed the contribution guidelines.
The PR follows the pull request guidelines.

In-review checklist for the PR reviewers

The PR reviewers must check the checkboxes below before approval.

The PR follows the pull request guidelines.
The PR has been properly tested.
The PR has been reviewed by the code owners.

Post-review checklist for the PR author

The PR author must check the checkboxes below before merging.

There are no open discussions or they are tracked via tickets.
The PR is ready for merge.

After all checkboxes are checked, anyone who has write access can merge the PR.

satoshi-ota · 2022-11-19T07:45:03Z

working on documentation update in #2331

satoshi-ota · 2022-11-22T00:57:47Z

@rej55 @TakaHoribe @zulfaqar-azmi-t4 @mkuri Could you review this and following PRs? 🙇

zulfaqar-azmi-t4 · 2022-11-25T05:55:39Z

...planning/lane_driving/behavior_planning/behavior_path_planner/avoidance/avoidance.param.yaml

@@ -15,6 +15,8 @@
      threshold_time_object_is_moving: 1.0        # [s]
      object_check_forward_distance: 150.0        # [m]
      object_check_backward_distance: 2.0         # [m]
+      object_check_shiftable_ratio: 0.6           # [-]
+      object_check_min_road_shoulder_width: 0.5   # [m]


I think these three related parameters should be rearranged, add white space on top and bottom so that it is simpler to be found

threshold_distance_object_is_on_center

shiftable ration

min road shoulder width

Currently threshold_distance_object_is_on_center is on top of the list.

Thanks for your advise 😄 I'll fix it.

I fixed in a94873d

TakaHoribe · 2022-11-29T00:44:55Z

assuming that the shoulder exists.

@satoshi-ota Would you explain the background/advantage of this assumption? Is this min_road_shoulder_length applied even if there is a sidewalk right next to the driving lane?

Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com>

…rocess Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com>

Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com>

codecov · 2022-11-29T09:19:44Z

Codecov Report

Base: 10.47% // Head: 10.47% // No change to project coverage 👍

Coverage data is based on head (cd01c88) compared to base (cd01c88).
Patch has no changes to coverable lines.

❗ Current head cd01c88 differs from pull request most recent head ccfa88a. Consider uploading reports for the commit ccfa88a to get more accurate results

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #2329   +/-   ##
=======================================
  Coverage   10.47%   10.47%           
=======================================
  Files        1248     1248           
  Lines       91307    91307           
  Branches    21035    21035           
=======================================
  Hits         9566     9566           
  Misses      71555    71555           
  Partials    10186    10186

Flag	Coverage Δ
total	`10.45% <0.00%> (ø)`

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

satoshi-ota · 2022-11-29T10:27:02Z

@TakaHoribe I think the left boundary of the road lanelet is not on acutual road bound (B) but on road outside line (A).

[Now, I am checking to make sure that is correct... ]

[TIER IV INTERNAL LINK]

P.S. Left bound of road lanelet is on road outside line (A) at least of odaiba map.

(Assuming I am correct) In many cases, vehicles can straddle the left boundary of lanelet and park on the left side of the road even if there is no explicit road shoulder lanelet. Therefore, shiftable ratio was calculated assuming that there is narrow (min_road_shoulder_length meter in width) road shoulder.

satoshi-ota · 2022-12-01T09:42:26Z

@TakaHoribe @zulfaqar-azmi-t4 @mkuri Do you have any other concern or unclear point or advice?

TakaHoribe

LGTM

rej55

LGTM

…2329) * feat(route_handler): add getMostLeftLanelet() Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * feat(avoidance): calc shiftable ratio in avoidance target filtering process Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * feat(avoidance): output object's debug info for rviz Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * fix(avoidance): use avoidance debug factor Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * feat(tier4_planning_launch): add new params for avoidance Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * fix(avoidance): reorder params for readability Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * fix(tier4_planning_launch): reorder params for readability Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> Signed-off-by: Kotaro Yoshimoto <pythagora.yoshimoto@gmail.com>

…2329) * feat(route_handler): add getMostLeftLanelet() Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * feat(avoidance): calc shiftable ratio in avoidance target filtering process Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * feat(avoidance): output object's debug info for rviz Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * fix(avoidance): use avoidance debug factor Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * feat(tier4_planning_launch): add new params for avoidance Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * fix(avoidance): reorder params for readability Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> * fix(tier4_planning_launch): reorder params for readability Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com> Signed-off-by: kminoda <koji.minoda@tier4.jp>

github-actions bot added component:launch Launch files, scripts and initialization tools. (auto-assigned) component:planning Route planning, decision-making, and navigation. (auto-assigned) labels Nov 19, 2022

satoshi-ota force-pushed the feat/improve-avoidance-target-filter branch from 30a6493 to 551cb36 Compare November 19, 2022 05:13

satoshi-ota mentioned this pull request Nov 19, 2022

feat(tier4_planning_msgs): improve avoidance target filter tier4/tier4_autoware_msgs#64

Merged

7 tasks

satoshi-ota marked this pull request as ready for review November 19, 2022 07:21

satoshi-ota requested review from rej55, a team, kosuke55, shmpwk, taikitanaka3, TakaHoribe, takayuki5168, tkimura4, purewater0901, zulfaqar-azmi-t4, kyoichi-sugahara, mkuri, soblin and TomohitoAndo as code owners November 19, 2022 07:21

satoshi-ota mentioned this pull request Nov 19, 2022

docs(avoidance): update avoidance documentation #2331

Closed

4 tasks

satoshi-ota mentioned this pull request Nov 20, 2022

feat(planning_launch): add new param to avoidance.param.yaml tier4/autoware_launch-old#569

Merged

8 tasks

satoshi-ota force-pushed the feat/improve-avoidance-target-filter branch from 551cb36 to eaa49cf Compare November 20, 2022 07:22

zulfaqar-azmi-t4 reviewed Nov 25, 2022

View reviewed changes

satoshi-ota requested a review from zulfaqar-azmi-t4 November 29, 2022 00:19

satoshi-ota added 3 commits November 29, 2022 17:36

feat(route_handler): add getMostLeftLanelet()

fa1b319

Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com>

feat(avoidance): calc shiftable ratio in avoidance target filtering p…

11905a1

…rocess Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com>

feat(avoidance): output object's debug info for rviz

9824ecc

Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com>

satoshi-ota added 4 commits November 29, 2022 17:36

fix(avoidance): use avoidance debug factor

7e5caa3

Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com>

feat(tier4_planning_launch): add new params for avoidance

49745a0

Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com>

fix(avoidance): reorder params for readability

7391a70

Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com>

fix(tier4_planning_launch): reorder params for readability

ccfa88a

Signed-off-by: satoshi-ota <satoshi.ota928@gmail.com>

satoshi-ota force-pushed the feat/improve-avoidance-target-filter branch from a94873d to ccfa88a Compare November 29, 2022 08:36

TakaHoribe approved these changes Dec 2, 2022

View reviewed changes

satoshi-ota enabled auto-merge (squash) December 2, 2022 00:31

rej55 approved these changes Dec 2, 2022

View reviewed changes

satoshi-ota merged commit cd729c5 into autowarefoundation:main Dec 2, 2022

tier4-autoware-public-bot bot mentioned this pull request Dec 3, 2022

chore: sync upstream tier4/autoware.universe#198

Merged

satoshi-ota mentioned this pull request Feb 27, 2023

docs(avoidance): update documentation of avoidance module #2954

Merged

4 tasks

satoshi-ota mentioned this pull request Mar 31, 2023

Avoidance module development plan #3230

Open

13 tasks

satoshi-ota deleted the feat/improve-avoidance-target-filter branch April 8, 2023 01:02

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(avoidance): improve avoidance target filter #2329

feat(avoidance): improve avoidance target filter #2329

satoshi-ota commented Nov 19, 2022 •

edited

Loading

satoshi-ota commented Nov 19, 2022

satoshi-ota commented Nov 22, 2022 •

edited

Loading

zulfaqar-azmi-t4 Nov 25, 2022

satoshi-ota Nov 25, 2022

satoshi-ota Nov 29, 2022

TakaHoribe commented Nov 29, 2022 •

edited

Loading

codecov bot commented Nov 29, 2022 •

edited

Loading

satoshi-ota commented Nov 29, 2022 •

edited

Loading

satoshi-ota commented Dec 1, 2022

TakaHoribe left a comment

rej55 left a comment

feat(avoidance): improve avoidance target filter #2329

feat(avoidance): improve avoidance target filter #2329

Conversation

satoshi-ota commented Nov 19, 2022 • edited Loading

Description

Judge whether the object is parking vehicle on road shoulder

Related links

Tests performed

Notes for reviewers

Pre-review checklist for the PR author

In-review checklist for the PR reviewers

Post-review checklist for the PR author

satoshi-ota commented Nov 19, 2022

satoshi-ota commented Nov 22, 2022 • edited Loading

zulfaqar-azmi-t4 Nov 25, 2022

Choose a reason for hiding this comment

satoshi-ota Nov 25, 2022

Choose a reason for hiding this comment

satoshi-ota Nov 29, 2022

Choose a reason for hiding this comment

TakaHoribe commented Nov 29, 2022 • edited Loading

codecov bot commented Nov 29, 2022 • edited Loading

Codecov Report

satoshi-ota commented Nov 29, 2022 • edited Loading

satoshi-ota commented Dec 1, 2022

TakaHoribe left a comment

Choose a reason for hiding this comment

rej55 left a comment

Choose a reason for hiding this comment

satoshi-ota commented Nov 19, 2022 •

edited

Loading

satoshi-ota commented Nov 22, 2022 •

edited

Loading

TakaHoribe commented Nov 29, 2022 •

edited

Loading

codecov bot commented Nov 29, 2022 •

edited

Loading

satoshi-ota commented Nov 29, 2022 •

edited

Loading