lu4r_ros_interface

This package enables for the communication between the LU4R system and ROS. This project is in charge of receiving the list of transcriptions from an external ASR and forward it to LU4R, and afterwards interpret each command to be able to move around a Pioneer 3-DX.

This ROS package provides tree different ROS nodes:

• audiolistener
• simlab_interface
• simlab_interpreter
• rosarnl_noderosarnl

In case of a simulation MobileSim will also be needed.

audiolistener

Listens to what the user has to say, and communicates with an external ASR to receive a transcript. This transcript is later used by the simlab_interface

simlab_interface

Receives a transcript and communicates with the LU4R server, it then receives a parsed command. It will be interpreted by the simlab_interpreter

Requirements

• ROS
• Python 2.7
• docker (Optional)

Run

Place the ROS package lu4r_ros_interface in your catkin_ws/src folder. Then, run

roscore

to start the ROS master. To start the LU4R server In another terminal, run:

java -jar -Xmx1G lu4r.jar [type] [output] [language] [port] #in lu4r server path

Or if you want to use docker, add the LU4R server documents into a folder with the Dockerfile found in /docker folder and run the following commands in a terminal:

cd [PATH_TO_LU4R_DOCKER_FOLDER]
docker build --tag=lu4r-server .
docker run -it --rm -p 9090:9090 lu4r-server

Now all you have to do is run the roslaunch file. If you are running a simulation, run the following command in a terminal:

MobileSim -m [PATH_TO_MAP]

And run the roslaunch in another terminal

roslaunch lu4r_ros_interface simlab_launch_simulated.launch 
lu4r_ip:=[lu4r_ip_address] lu4r_port:=[lu4r_port] semantic_map:=[semantic_map]

where:

• \_lu4r\_ip\_address: the ip address of LU4R server. If the LU4R and the LU4R ROS interface are on the same machine, ignore this argument, as it is set to 127.0.0.1 by default.

• \_lu4r\_port: the listening port of LU4R server. It is set to 9090 by default.

• \_semantic\_map: the semantic map to be used, among the ones available into semantic_maps. Semantic maps are in JSON format, representing the configuration of the environment (e.g. objects, locations,...) in which the robot is operating. Whenever a simple configuration of LU4R is chosen, the interpretation process is sensitive to different semantic maps. By default "semantic_map1.txt".

---

If it's going to be runned in a Pioneer 3-DX robot, you have to make the robot run the following in serveral terminals:

roscore

and

rosrun rosarnl rosarnl_node

Make sure that the host where the other nodes will be running has their ROS variables correctly definded. Run these commands to set them:

export ROS_IP=XXX.XXX.XXX.XXX # our IP
export ROS_HOSTNAME=XXX.XXX.XXX.XXX # our IP
export ROS_MASTER_URI=http://YYY.YYY.YYY.YYY:PORT # the remote machine IP (PORT by default is 11311, but it depends on the robot's configuration you use).

Once everything has beed configured, run

roslaunch lu4r_ros_interface simlab_launch.launch 
lu4r_ip:=[lu4r_ip_address] lu4r_port:=[lu4r_port] semantic_map:=[semantic_map]

For more information, please visit the LU4R system website, or take a look to the documentation found in the docs folder of this repository.