PredictionLayer

---

contents

1. Concepts
2. Getting Started
3. Parameters
4. Example guide
5. ToDo

1. Concepts

---

This package is only for ros-melodic user.

using obstacle_detector package for perception function, So you have to use 2D Lidar and obstacle_detector package.

2. Getting Started

---

install guide before getting started

1. install source from github and install dependency packages

   1. [obstacle_detector|!https://github.com/tysik/obstacle_detector.git] is default reference dependency package for this layer, so you have to install this package first

2. After, install obstacle_detector package, follow the guide below

   1. install git source from github

      cd ~/usr_ws/src
      git clone https://github.com/OkDoky/prediction_layer.git
      cd ~/usr_ws

   2. install dependency packages

      rosdep install -r -y --from-paths src --ignore-src
      catkin_make

3. Parameters

---

• ~enabled (bool, default : true) - enable/disable layer function
• ~footprint_clearing_enabled (bool, default : true) - enable/disable clear costs inside robot footprint
• ~object_source (string, default : "") - topic name for input data, the input data type is obstacle_detector/Obstacles, default name is obstacles
• ~observation_persistence (double, default : 0.2) -
• ~expected_update_rate (double, default : 0.2) -
• ~clearing (bool, default : true) - enable/disable clear the marked buffer
• ~marking (bool, defualt : true) -
• ~transform_tolerance (double, default : 0.2) -
• ~track_unknown_space (bool, default : true) -
• ~combination_method (int, default : 0) -

4. Example guide

---

example for costmap layer

1. example guide

5. ToDo

---

• non-buffer layer
• reduce cycle time using reactiveX (rxcpp)
• refactoring