Event-Driven Manufacturing - Basics
Many are the articles describing Event-Driven Manufacturing concepts, but I think that it is worth to attain to the basics to facilitate the understanding of the concept and further realize how to implement it in the real production world.
The most noticeable fact is that events happens all the time in any day-to day manufacturing operation, if you notice well the production starts, operators interfere opening machine doors to check working conditions, machine stops or raises alerts about faulty conditions or warnings, operation and process parameters goes out of boundaries due to lack of machine maintenance or operating errors, pieces have identifiers or data matrix code engraved on them for posterior tracking and finally the production stops and the piece is finally unloaded from the machine, the sequence of events tracking continues up to the point that the piece leaves the facility or there is no need to track the events anymore.
Whether the produced piece is good or bad it depends on lots or factors or events, but better will be if we can detect the piece quality as good or bad as early as possible to avoid further work or waste of work in case it is a possible scrap. In a traditional control system, which is not event-driven the error detection can be done at a late stage when lots of effort and time-consuming operations might be already executed on the scrapped part.
In a well-designed Event-Driven Manufacturing system all production events are tracked and monitored and in case parameter values are out the allowed boundaries, notifications are immediately sent to the interested persons, to other systems, machines or departments to allow that specific actions be taken in a timely fashion.
Modern technologies not by coincidence are designed to facilitate the implementation of such systems, but the traditional paradigms and change resistance sometimes play an important factor to delay this adoption. A good example of such facilitating technologies is OPC-UA with its Pub/Sub concept, MQTT, Enterprise message buses such as Apache Kafka and others, MS SQL Broker, Enterprise Service Buses, SOA, ETL, REST APIs, Microservices Architecture, Kubernetes, and the whole set of existent cloud services.
All those are truly event-driven enabler technologies yet just a few or not at all explored in many of the existent IT/OT manufacturing execution systems. One must think and say that this is so new and unknown, and ask himself if it does really work; So better is if we ask to Netflix, LinkedIn, Twitter, and many other younger companies that uses these technologies very successfully. Apart from other systems, manufacturing related systems such as SCADA, MES, QMS, PLM, ERP are basically and traditionally transactional systems and many are not yet prepared to deal with events in a seamless manner.
So how to implement this concept on IT Manufacturing systems; It is time to start the proof-of-concept projects and explore the cloud technologies to speed up your test projects and adoption, as happen in the real world, we normally do not rebuild a house from scratch, it is easy if we add other rooms to the existent ones in order to fulfill our current needs…
Recommended by LinkedIn
By using cloud modules such as Azure IoT Edge, IoT Hub, OPC Publisher and others one can quickly and easily do a proof-of-concept using all those mentioned enablers which are indeed cloud agnostic; Similar modules are also found in other cloud providers such as AWS, Google, IBM, and others.
Below you can find an overall diagram of a possible architecture considering cloud components and machines powered with Windows Based HMIs:
On the PLC and On-Premises side we can have something like this, here I considered a finite state machine (FSM) to make the PLC Engineer task of mapping the machine signals to defined events, also thinking in the standardization aspects the usage of an OPC-UA companion specification designed to standardize the events of the specific processes comes in handy.
If you are also interested in this topic, please share your opinions, ideas and comments here, I hope you have enjoyed the reading!