This project contains the implementation of a proxemic layer for the costmaps of the ROS navigation stack.
proxemic_layer is a layer for the costmap generation of the ROS navigation stack. A detailed description can be found in this paper.
@inproceedings{
title = {Modelling Proxemics for Human-Technology-Interaction in Decentralized Social-Robot-Systems},
author = {Thomas Kirks, Jana Jost, Jan Finke, Sebastian Hoose},
year = {2020},
booktitle = {Intelligent Human Systems Integration 2020},
publisher = {Springer}
pages = {153 - 158}
}
people_publisher is a node which publishes test data for the proxemic_layer. It simply provides a dynamic_reconfigure parameterset which, after setting e.g. using rqt, is published to the /people topic, which again, is subscribed by the proxemic_layer.
Simply copy the proxemic_layer into your ROS workspace. After doing so, you can start using the proxemic_layer as a layer of the costmap. See below for an example config file. The people_publishing is only provided for testing purpose, it is not necessary for the proxemic_layer to work.
- Usage details for the proxemic_layer
- Usage details for the people_publisher
plugins:
- {name: static_map, type: "costmap_2d::StaticLayer"}
- {name: obstacles, type: "costmap_2d::VoxelLayer"}
- {name: proxemic_layer, type: "proxemic_layer::ProxemicLayer"}
static_map:
lethal_cost_threshold: 55
obstacles:
#obstacles from sensors definition
proxemic_layer:
enabled: true
max_time_passed: 10
gaussian_renorming: 150
An example of the resulting proxemic: