Detection Area

Role

If pointcloud or predicted objects are detected in a detection area defined on a map, the stop planning will be executed at the predetermined point.

Activation Timing

This module is activated when there is a detection area on the target lane.

Module Parameters

Inner-workings / Algorithm

1. Gets a detection area and stop line from map information and confirms if there are obstacles in the detection area
2. Inserts stop point l[m] in front of the stop line
3. Calculates required braking distance based on current velocity, max_deceleration, and delay_response_time
4. If the vehicle cannot stop before the stop line, applies the configured unstoppable_policy
5. Sets velocity as zero at the determined stop point

Detection Logic

The module uses two detection sources:

• Pointcloud detection: Detects any obstacles in the detection area using 3D point cloud data (if target_filtering.pointcloud is enabled)
• Predicted objects detection: Detects classified objects (vehicles, pedestrians, etc.) in the detection area based on perception module outputs

The module stops the vehicle if either detection source finds an obstacle. For performance optimization, if pointcloud detection finds an obstacle, predicted objects detection is skipped (short-circuit evaluation).

Flowchart

```plantuml @startuml title modifyPathVelocity start

:check obstacles area;

if (obstacle detected?) then (no) else (yes) :set last time obstacle found; endif

:get clear stop state duration;

if (clear stop state duration is more than state_clear_time?) then (yes) if (suppress_pass_judge_when_stopping is false or ego is moving) then (yes) :set current state GO; endif :reset clear stop state duration; stop else (no) endif

if (use dead line?) then (yes) :create dead line point; if (Is there dead line point?) then (yes) if (Is over dead line point?) then (yes) stop endif endif endif

:calculate stop point;

if (stop point?) then (yes) else (no) stop endif

File truncated at 100 lines see the full file

Changelog for package autoware_behavior_velocity_detection_area_module

0.48.0 (2025-11-18)

• Merge remote-tracking branch 'origin/main' into humble

• feat(detection_area): implement unified handling for unstoppable situations (#11638)

  * feat(detection_area): implement unified handling for unstoppable situations Implement unified handling for unstoppable situations in both detection_area and detection_area/experimental modules with enhanced policy handling.

  • fix(detection_area): update unstoppable_policy to stop_after_stopline

  * docs(detection_area): enhance unstoppable situation handling with new parameters ---------

• feat(detection_area): update object detection logic to use polygon intersection for detection areas (#11625)

• feat(behavior_velocity_detection_area_module): add target filtering parameters to DetectionAreaModuleManager (#11636)

• feat(behavior_velocity_detection_area_module): use Trajectory instead of PathWithLaneId (#11583)

  • use Trajectory instead of PathWithLaneId

  * Update planning/behavior_velocity_planner/autoware_behavior_velocity_detection_area_module/src/experimental/scene.cpp Co-authored-by: Mamoru Sobue <<mamoru.sobue@tier4.jp>>

  * Update planning/behavior_velocity_planner/autoware_behavior_velocity_detection_area_module/src/experimental/scene.cpp ---------Co-authored-by: Yukinari Hisaki <<42021302+yhisaki@users.noreply.github.com>> Co-authored-by: Mamoru Sobue <<mamoru.sobue@tier4.jp>>

• fix: tf2 uses hpp headers in rolling (and is backported) (#11620)

• feat(behavior_velocity_rtc_interface, behavior_velocity_*_module): replace PathWithLaneId with Trajectory<> class (#11555)

• feat(detection_area): enable stopping based on predicted objects (#11588)

  • feat(detection_area): enhance obstacle detection with target filtering for various object types
  • Added support for detecting predicted objects in the detection area alongside pointcloud data.
  • Introduced target filtering parameters in the configuration to specify which object types (e.g., cars, trucks, pedestrians) should trigger stop planning.
  • Updated README to reflect changes in detection logic and configuration options.

  - Modified relevant source files to implement the new detection logic and parameter handling. This enhancement improves the module's ability to respond to a wider range of obstacles, increasing safety and reliability in planning scenarios.

  • docs(detection_area): update README to clarify detection area and target filtering parameters

  * refactor(detection_area): simplify detection source assignment and enhance target object filtering ---------

• chore(detection_area): add Yukinari Hisaki and Takumi Odashima as maintainers (#11587) chore(package.xml): add Yukinari Hisaki and Takumi Odashima as maintainers

• Contributors: Mitsuhiro Sakamoto, Ryohsuke Mitsudome, Tim Clephas, Yukinari Hisaki

0.47.1 (2025-08-14)

0.47.0 (2025-08-11)

• style(pre-commit): update to clang-format-20 (#11088) Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

• feat(detection_area): set stop point behind actual stop line when ego exceeds max acceleration (#10919)

  • feat(detection_area): set stop point behind actual stop line

File truncated at 100 lines see the full file