Repository : https://gitlab.fit.fraunhofer.de/efpf-pilots/roam
Q: Where can I find the ROAM tool? A: http://efpf.almende.com
Q: What are “recipes”? A: Recipes are transformations that turn some input into some output. For example, a recipe can output the average of the input data, or it can send an email notification if the input data exceeds a certain threshold while outputting the input data unchanged.
Q: What are “workflows”? A: Workflows are recipes glued together, along with a defined input and output MQTT topic on the EFPF data spine broker. Whenever there’s a message on the input topic, it is passed to the first recipe in the workflow. Each recipe then sequentially converts its input to an output, and that output is passed to the next recipe as input. The final output is published to the output topic.
Q: What does using the ROAM tool cost? A: The ROAM tool can be used for free, which allows users to create public base recipes (which are free to use for all EFPF users). Configured recipes and workflows are private and are only shared with whoever the user authorizes to see them. Private base recipes can be submitted when a user or company is subscribed to the ROAM tool. However, we are allowed to gather anonymized data on used recipes and parameters; the user is allowed to hide a limited amount of parameters, when they are confidential or possibly too revealing. The users can also opt to pay per recipe to make them private.
Q: I don’t know how to write a base recipe, but need something specific. What now? Almende offers a consulting service to define and develop recipes. We can also help identify risks and insightful metrics to monitor.
Q: What benefits does the ROAM tool provide? A: The ROAM tool is integrated with the EFPF architecture and is directly connected to its MQTT broker. The tool allows users obtain real time metrics and risks from their connected machines. To name some examples: The user can extrapolate their current production to determine whether they will finish production in time for their order. The user can use sensor data to determine whether their machines are likely to need maintenance or are running too hot. The user can receive notifications whenever specific criteria are met, such as those in the previous points