Skip to content

Commit 6a80b3b

Browse files
brkay54brkay54
committed
feat(reaction_analyzer): add reaction anaylzer tool to measure end-to-end delay in sudden obstacle braking response
Signed-off-by: Berkay Karaman <brkay54@gmail.com>
1 parent ebb4172 commit 6a80b3b

11 files changed

+3246
-0
lines changed

‎tools/reaction_analyzer/CMakeLists.txt

+39
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,39 @@
1+
cmake_minimum_required(VERSION 3.14)
2+
project(reaction_analyzer)
3+
4+
find_package(autoware_cmake REQUIRED)
5+
autoware_package()
6+
7+
find_package(PCL REQUIRED)
8+
find_package(Eigen3 REQUIRED)
9+
10+
ament_auto_add_library(reaction_analyzer SHARED
11+
include/reaction_analyzer_node.hpp
12+
src/reaction_analyzer_node.cpp
13+
include/topic_publisher.hpp
14+
src/topic_publisher.cpp
15+
)
16+
17+
target_include_directories(reaction_analyzer
18+
SYSTEM PUBLIC
19+
${PCL_INCLUDE_DIRS}
20+
${EIGEN3_INCLUDE_DIR}
21+
)
22+
23+
target_link_libraries(reaction_analyzer
24+
${PCL_LIBRARIES}
25+
)
26+
27+
rclcpp_components_register_node(reaction_analyzer
28+
PLUGIN "reaction_analyzer::ReactionAnalyzerNode"
29+
EXECUTABLE reaction_analyzer_exe
30+
)
31+
32+
if (BUILD_TESTING)
33+
endif ()
34+
35+
ament_auto_package(
36+
INSTALL_TO_SHARE
37+
param
38+
launch
39+
)

‎tools/reaction_analyzer/README.md

+137
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,137 @@
1+
# Reaction Analyzer
2+
3+
## Description
4+
5+
The main purpose of the reaction analyzer package is measuring reaction times of the nodes by listening the
6+
pre-determined topics. It can be used for measuring the reaction time of the perception pipeline, planning pipeline, and
7+
control pipeline. To be able to measure both control outputs and perception outputs, it was necessary to divide the node
8+
into two running_mode: `planning_control` and `perception_planning`.
9+
10+
![ReactionAnalyzerDesign.png](media%2FReactionAnalyzerDesign.png)
11+
12+
### Planning Control Mode
13+
14+
In this mode, the reaction analyzer creates a dummy publisher for the PredictedObjects and PointCloud2 topics. In the
15+
beginning of the test, it publishes the initial position of the ego vehicle and the goal position to set the test
16+
environment. Then, it spawns a sudden obstacle in front of the ego vehicle. After the obstacle is spawned, it starts to
17+
search reacted messages of the planning and control nodes in the pre-determined topics. When all the topics are reacted,
18+
it calculates the reaction time of the nodes and statistics by comparing `reacted_times` of each of the nodes
19+
with `spawn_cmd_time`, and it creates a csv file to store the results.
20+
21+
### Perception Planning Mode
22+
23+
In this mode, the reaction analyzer reads the rosbag files which are recorded from AWSIM, and it creates a topic
24+
publisher for each topic inside the rosbag to replay the rosbag. It reads two rosbag files: `path_bag_without_object`
25+
and `path_bag_with_object`. Firstly, it replays the `path_bag_without_object` to set the initial position of the ego
26+
vehicle and the goal position. After `spawn_time_after_init` seconds , it replays the `path_bag_with_object` to spawn a
27+
sudden obstacle in front of the ego vehicle. After the obstacle is spawned, it starts to search the reacted messages of
28+
the perception and planning nodes in the pre-determined topics. When all the topics are reacted, it calculates the
29+
reaction time of the nodes and statistics by comparing `reacted_times` of each of the nodes with `spawn_cmd_time`, and
30+
it creates a csv file to store the results.
31+
32+
#### Point Cloud Publisher Type
33+
34+
To get better analyze for Perception & Sensing pipeline, the reaction analyzer can publish the point cloud messages in 3
35+
different ways: `async_header_sync_publish`, `sync_header_sync_publish` or `async_publish`. (`T` is the period of the
36+
lidar's output)
37+
38+
![PointcloudPublisherType.png](media%2FPointcloudPublisherType.png)
39+
40+
- `async_header_sync_publish`: It publishes the point cloud messages synchronously with asynchronous header times. It
41+
means that each of the lidar's output will be published at the same time, but the headers of the point cloud messages
42+
includes different timestamps because of the phase difference.
43+
- `sync_header_sync_publish`: It publishes the point cloud messages synchronously with synchronous header times. It
44+
means that each of the lidar's output will be published at the same time, and the headers of the point cloud messages
45+
includes the same timestamps.
46+
- `async_publish`: It publishes the point cloud messages asynchronously. It means that each of the lidar's output will
47+
be published at different times.
48+
49+
## Usage
50+
51+
The common parameters you need to define for both running modes are `output_file_path`, `test_iteration`,
52+
and `reaction_chain` list. `output_file_path` is the output file path is the path where the results and statistics
53+
will be stored. `test_iteration` defines how many tests will be performed. The `reaction_chain` list is the list of the
54+
pre-defined topics you want to measure their reaction times.
55+
56+
### Prepared Test Environment
57+
58+
Firstly, you need to download the demonstration test map from the
59+
link [here](https://github.com/tier4/AWSIM/releases/download/v1.1.0/nishishinjuku_autoware_map.zip). After downloading,
60+
extract the zip file and use its path as `[MAP_PATH]` in the following commands.
61+
62+
#### Planning Control Mode
63+
64+
Firstly, you need to define only Planning and Control nodes in the `reaction_chain` list. With the default parameters,
65+
you can start to test with the following command:
66+
67+
```bash
68+
ros2 launch reaction_analyzer reaction_analyzer.launch.xml running_mode:=planning_control vehicle_model:=sample_vehicle sensor_model:=sample_sensor_kit map_path:=[MAP_PATH]
69+
```
70+
71+
After the command, the simple_planning_simulator and the reaction_analyzer will be launched. It will automatically start
72+
to test. After the test is completed, the results will be stored in the `output_file_path` you defined.
73+
74+
#### Perception Planning Mode
75+
76+
Firstly, you need to download the rosbag files from the
77+
link [here](https://drive.google.com/file/d/1_P-3oy_M6eJ7fk8h5CP8V6s6da6pPrBu/view?usp=sharing). After downloading,
78+
extract the zip file and set the path of the `.db3` files to parameters `path_bag_without_object`
79+
and `path_bag_with_object`. Because custom sensor setup, you need to checkout the following branches before launch the
80+
reaction analyzer: For the `autoware_individual_params` repository, checkout the
81+
branch [here](https://github.com/brkay54/autoware_individual_params/tree/bk/reaction-analyzer-config). For
82+
the `awsim_sensor_kit_launch` repository, checkout the
83+
branch [here](https://github.com/brkay54/awsim_sensor_kit_launch/tree/bk/reaction-analyzer-config).
84+
After you checkouted the branches, you can start to test with the following command:
85+
86+
```bash
87+
ros2 launch reaction_analyzer reaction_analyzer.launch.xml running_mode:=perception_planning vehicle_model:=sample_vehicle sensor_model:=awsim_sensor_kit map_path:=[MAP_PATH]
88+
```
89+
90+
After the command, the e2e_simulator and the reaction_analyzer will be launched. It will automatically start
91+
to test. After the test is completed, the results will be stored in the `output_file_path` you defined.
92+
93+
### Custom Test Environment
94+
95+
If you want to run the reaction analyzer with your custom test environment, you need to redefine some of the parameters.
96+
The parameters you need to redefine are `initialization_pose`, `entity_params`, `goal_pose`, and `topic_publisher`
97+
parameters.
98+
For `initialization_pose`, `entity_params`, `goal_pose`, you can upload your `.osm` map file into
99+
the [scenario editor](https://scenario.ci.tier4.jp/scenario_editor/) to define the position of the position parameters.
100+
You can define the positions with respect to you desired root. Firstly, add EGO vehicle from edit/add entity/Ego to map.
101+
Set destination to EGO vehicle and add another dummy object in same way. The dummy object represents the object spawn
102+
suddenly in the reaction analyzer test. After you set up the positions in the map, we should get the positions of these
103+
entities in the map frame. To achieve this, you should convert the positions to map frame by changing Map/Coordinate to
104+
World and Map/Orientation to Euler. After these steps, you can see the positions in map frame and euler angles. You can
105+
change the `initialization_pose`, `entity_params`, `goal_pose` parameters with the values you get from the website.
106+
107+
For the `topic_publisher` parameters, you need to record the rosbags from the AWSIM. After opened your AWSIM
108+
environment, you should record two different rosbags. However, the environment should be static and the position of the
109+
vehicle should be same. Firstly, record a rosbag in empty environment. After that, record another rosbag in the same
110+
environment except add an object in front of the vehicle. After you record the rosbags, you can set
111+
the `path_bag_without_object` and `path_bag_with_object` parameters with the paths of the recorded rosbags.
112+
113+
## Parameters
114+
115+
| Name | Type | Description |
116+
| ------------------------------------------------------------------ | ------ | --------------------------------------------------------------------------------------------------------------------------------------------- |
117+
| `timer_period` | double | [s] Period for the main processing timer. |
118+
| `published_time_expire_duration` | double | [s] Time horizon of the PublishedTime message vector. |
119+
| `test_iteration` | int | Number of iterations for the test. |
120+
| `output_file_path` | string | Directory path where test results and statictics will be stored. |
121+
| `object_search_radius_offset` | double | [m] Additional radius added to the search area when looking for objects. |
122+
| `spawn_time_after_init` | double | [s] Time delay after initialization before spawning objects. Only valid `perception_planning` mode. |
123+
| `spawn_distance_threshold` | double | [m] Distance threshold for spawning objects. Only valid `planning_control` mode. |
124+
| `spawned_pointcloud_sampling_distance` | double | [m] Sampling distance for point clouds of spawned objects. Only valid `planning_control` mode. |
125+
| `dummy_perception_publisher_period` | double | [s] Publishing period for the dummy perception data. Only valid `planning_control` mode. |
126+
| `first_brake_params.debug_control_commands` | bool | Debug publish flag. |
127+
| `first_brake_params.control_cmd_buffer_time_interval` | double | [s] Time interval for buffering control commands. |
128+
| `first_brake_params.min_number_descending_order_control_cmd` | int | Minimum number of control commands in descending order for triggering brake. |
129+
| `first_brake_params.min_jerk_for_brake_cmd` | double | [m/s³] Minimum jerk value for issuing a brake command. |
130+
| `initialization_pose` | struct | Initial pose of the vehicle, containing `x`, `y`, `z`, `roll`, `pitch`, and `yaw` fields. Only valid `planning_control` mode. |
131+
| `entity_params` | struct | Parameters for entities (e.g., obstacles), containing `x`, `y`, `z`, `roll`, `pitch`, `yaw`, `x_dimension`, `y_dimension`, and `z_dimension`. |
132+
| `goal_pose` | struct | Goal pose of the vehicle, containing `x`, `y`, `z`, `roll`, `pitch`, and `yaw` fields. |
133+
| `topic_publisher.path_bag_without_object` | string | Path to the ROS bag file without objects. Only valid `perception_planning` mode. |
134+
| `topic_publisher.path_bag_with_object` | string | Path to the ROS bag file with objects. Only valid `perception_planning` mode. |
135+
| `topic_publisher.pointcloud_publisher.pointcloud_publisher_type` | string | Defines how the PointCloud2 messages are going to be published. Modes explained above. |
136+
| `topic_publisher.pointcloud_publisher.pointcloud_publisher_period` | double | [s] Publishing period of the PointCloud2 messages. |
137+
| `reaction_chain` | struct | List of the nodes with their topics and topic's message types. |

0 commit comments

Comments
 (0)