This repository contains the implementation of the augmented (retrofitted) production and storage processes for the Fischertechnik Learning Factory Industry 4.0 (9V). The implementation accompanies the paper "A Process to Non-invasively Augment Legacy IoT Systems Using Business Processes and Microservices" published at the 14th International Conference on the Internet of Things (IoT 2024).
Here you can find 5 Java-based microservices implemented using Spring Boot to extend the existing processes in the learning factory with additional functionality. The Camunda microservice contains an embedded business process management system, namely Camunda Platform 7 (Community Edition), that manages, executes, and monitors the business processes.
We assume the standard, out-of-the-box setup of the Fischertechnik Learning Factory Industry 4.0 (9V) to be used, with all 5 TXT controllers connected to a local WiFi network with IP configurations according to this documentation. The computer used to run the software provided in this repository also needs to be connected to this network. Make sure that this computer is able to reach the smart factory's MQTT broker (per default, running at: 192.168.0.10:1883, user: txt, pwd: xtx).
Note that you can adjust parameters, including the MQTT broker configurations in the application.properties files of the individual microservices.
The implementation uses Maven as build system and Docker as runtime platform. Java 17 is used. All microservices are standalone Spring Boot applications.
First, all microservices need to be built by running:
mvn clean installOnce this process finished successfully. The Docker containers have to be built and started:
docker compose up --buildAll 5 microservices should be started in Docker containers afterward and the microservices should be connected to the factory's MQTT broker. The user interface of the BPMS will be available at: http://127.0.0.1:8080 (user: demo / pwd: demo) of the computer running this software project.
The two augmented business processes will be deployed automatically upon start. They can be found in folder smarter-factory/camunda/src/main/resources/ and modified using Camunda Modeler. New business processes can be modeled and uploaded to the BPMS using Camunda Modeler (see instructions).
Note that if there is no physical learning factory available. You can also simulate the execution of the original storage and production processes in the smart factory using a publicly available MQTT broker: The data folder in this repository contains exemplary data recorded from the original process execution in the smart factory via its MQTT interface. It can be replayed line by line, by sending these JSON messages, line by line, to the MQTT topics specified in the first attribute of each line as key. The actual payload of the message is the associated value. By paying attention to the embedded timestamps, the real-world execution can be simulated. The MQTT configuration of all microservices need to be adjusted according to the MQTT broker used (see application.properties files).