Oscar

This project is part of DIH^2. For more information check the RAMP Catalogue entry for the components.

📚 Documentation	🐳 Docker Hub	🎥 Video Tutorial	💾 OSCAR Dataset

Contents

• Background
• Requirements
• Instructions
• Testing
• Notes
• License

Background

Oscar is an IoT application for enabling communication between an interactive Graphical User Interface (GUI) and a robotic arm assisting the operator during the assembly process.

Main modules

1. CSV Parser: component in charge of loading the data of the components and processing steps on Fiware.
2. ROS Services for Fiware-UI-Robot: component enabling communication from UI to the robotic arm
3. ROS Services for Kinect-Fiware: component enabling communication from the camera to Fiware
4. HoloLens UI application: interactive Graphical User Interface

Requirements

CSV Data Format

Files containing data must be in csv format, using the comma , delimiter. As for now csv data is split among the following files:

Component.csv
Step.csv
Macrostep.csv
UseCase.csv

where the entity UseCase contains one or more entities Macrostep, which, in turn contain one or more entities Step.

• The first row of the csv tables contains the attribute name (first letter in lowercase) and in the case of relationship attributes their name convention must follow the rule: name_of_attribute = 'ref' + name_of_external_entity. So if, for example, an entity Step has an attribute linking to the Component entity the name of the attribute must be refComponent.

• The second row must contain the type of the attribute, the type of the first column containing the id must be of type None, while the relationship attributes must be of type Relationship

Instructions

To access the video tutorial showing all following steps please follow this link.

Only the first time execute the command:

docker network create fiware_default

then run the containers for mongo and orion:

docker run -d --name=mongo-db --network=fiware_default   --expose=27017 mongo:4.2 --bind_ip_all
docker run -d --name fiware-orion -h orion --network=fiware_default   -p 1026:1026  fiware/orion -dbhost mongo-db

To run the container execute the command

docker run --ipc=host  -it  phm14/oscar

Make sure that in the file config.yaml inside the folder src/oscar_core the entities address field entities_address: http://IP_ADDRESS:PORT_NUMBER/v2/entities/ is correct (for example if running on Linux on localhost use the address 172.17.0.1):

nano src/oscar_core/config.yml

and change accordingly entities_address: http://IP_ADDRESS:PORT_NUMBER/v2/entities/. Remember to check the robot ip address is correct.

Inside the container run the following:

source install/setup.bash
ros2 launch bring_up bring_up_launch.py

Load data

To load data on Fiware execute the command:

cd Scripts
python parser_to_fiware.py

the script file must be on the same folder of config.yml file to work

Testing

Inside the container run the following:

source install/setup.bash
ros2 launch parser parser_launch.py

To check if services are running launch another session from the same container:

docker exec -it <container_id> bash

inside the new session source ros:`

source install/setup.bash

To visualize data launch the command:

ros2 run parser ficlient TYPE ID

where TYPE can be one of the following values component, step, macrostep, usecase depending on the entity and ID is the same value found on the input csv files.

Notes

When launching the command ros2 launch bring_up bring_up_launch.py the services managing the robot are executed, to work the robot must be connected to the host.
An example of data gathered during a run of the application can be accessed on the following link.

License

License © 2023