diff --git a/planning/behavior_path_dynamic_avoidance_module/README.md b/planning/behavior_path_dynamic_avoidance_module/README.md index ab299297db94a..b8af767afd9a5 100644 --- a/planning/behavior_path_dynamic_avoidance_module/README.md +++ b/planning/behavior_path_dynamic_avoidance_module/README.md @@ -1,7 +1,9 @@ # Dynamic avoidance design + This module is under development. ## Purpose / Role + This module provides avoidance functions for vehicles, pedestrians, and obstacles in the vicinity of the ego's path in combination with the [obstacle_avoidance_planner](https://autowarefoundation.github.io/autoware.universe/main/planning/obstacle_avoidance_planner/). Each module performs the following roles. Dynamic Avoidance module cuts off the drivable area according to the position and velocity of the target to be avoided. @@ -19,10 +21,12 @@ The table below lists the avoidance modules that can handle each situation. | avoid moving objects | Dynamic Avoidance Module + Obstacle Avoidance Module | No Module (Under Development) | ## Policy of algorithms + Here, we describe the policy of inner algorithms. The inner algorithms can be separated into two parts: The first decides whether to avoid the obstacles and the second cuts off the drivable area against the corresponding obstacle. ### Select obstacles to avoid + To decide whether to avoid an object, both the predicted path and the state (pose and twist) of each object are used. The type of objects the user wants this module to avoid is also required. Using this information, the module decides to _avoid_ objects that _obstruct the ego's passage_ and _can be avoided_. @@ -33,12 +37,14 @@ For this purpose, the module judges whether the obstacle can be avoided with sat For example, the module decides not to avoid an object that is too close or fast in the lateral direction. ### Cuts off the drivable area against the selected vehicles + For the selected obstacles to be avoided, the module cuts off the drivable area. As inputs to decide the shapes of cut-off polygons, poses of the obstacles are mainly used, assuming they move in parallel to the ego's path, instead of its predicted path. This design arises from that the predicted path of objects is not accurate enough to use the path modifications (at least currently). Furthermore, the output drivable area shape is designed as a rectangular cutout along the ego's path to make the computation scalar rather than planar. #### Determination of lateral dimension + The lateral dimensions of the polygon are calculated as follows. The polygon's width to extract from the drivable area is the obstacle width and `drivable_area_generation.lat_offset_from_obstacle`. We can limit the lateral shift length by `drivable_area_generation.max_lat_offset_to_avoid`. @@ -46,6 +52,7 @@ We can limit the lateral shift length by `drivable_area_generation.max_lat_offse ![drivable_area_extraction_width](./image/drivable_area_extraction_width.drawio.svg) #### Determination of longitudinal dimension + Then, extracting the same directional and opposite directional obstacles from the drivable area will work as follows considering TTC (time to collision). Regarding the same directional obstacles, obstacles whose TTC is negative will be ignored (e.g., The obstacle is in front of the ego, and the obstacle's velocity is larger than the ego's velocity.). @@ -57,9 +64,11 @@ Opposite directional obstacles (Parameter names may differ from implementation) ![opposite_directional_object](./image/opposite_directional_object.svg) ### Cuts off the drivable area against the selected pedestrians + Then, we describe the logic to generate the drivable areas against pedestrians to be avoided. Objects of this type are considered to have priority right of way over the ego's vehicle while ensuring a minimum safety of the ego's vehicle. In other words, the module assigns a drivable area to an obstacle with a specific margin based on the predicted paths with specific confidences for a specific time interval, as shown in the following figure. +
Restriction areas are generated from each pedestrian's predicted paths
@@ -68,12 +77,12 @@ In other words, the module assigns a drivable area to an obstacle with a specifi Apart from polygons for objects, the module also generates another polygon to ensure the ego's safety, i.e., to avoid abrupt steering or significant changes from the path. This is similar to avoidance against the vehicles and takes precedence over keeping a safe distance from the object to be avoided. As a result, as shown in the figure below, the polygons around the objects reduced by the secured polygon of the ego are subtracted from the ego's drivable area. +
Ego's minimum requirements are prioritized against object margin
- ## Example
diff --git a/planning/behavior_path_dynamic_avoidance_module/include/behavior_path_dynamic_avoidance_module/scene.hpp b/planning/behavior_path_dynamic_avoidance_module/include/behavior_path_dynamic_avoidance_module/scene.hpp index 220286383fcfa..353a5fab92032 100644 --- a/planning/behavior_path_dynamic_avoidance_module/include/behavior_path_dynamic_avoidance_module/scene.hpp +++ b/planning/behavior_path_dynamic_avoidance_module/include/behavior_path_dynamic_avoidance_module/scene.hpp @@ -69,7 +69,7 @@ struct MinMaxValue ret.max_value = max_value + scalar; return ret; }; - void swap() { std::swap(min_value, max_value); }; + void swap() { std::swap(min_value, max_value); } }; enum class PolygonGenerationMethod { diff --git a/planning/behavior_path_dynamic_avoidance_module/src/scene.cpp b/planning/behavior_path_dynamic_avoidance_module/src/scene.cpp index 49b6b0eb89617..83db0f9edcc1e 100644 --- a/planning/behavior_path_dynamic_avoidance_module/src/scene.cpp +++ b/planning/behavior_path_dynamic_avoidance_module/src/scene.cpp @@ -1661,7 +1661,7 @@ DynamicAvoidanceModule::calcEgoPathBasedDynamicObstaclePolygon( return obj_poly; } -// TODO (takagi): replace by the function calcPredictedPathBasedDynamicObstaclePolygon()? +// should be replace by the function calcPredictedPathBasedDynamicObstaclePolygon() (takagi) std::optional DynamicAvoidanceModule::calcObjectPathBasedDynamicObstaclePolygon( const DynamicAvoidanceObject & object) const