Many life sciences manufacturers want to connect their manufacturing software to data sources such as equipment, devices, and third-party systems. In the past, such integration work has caused considerable project troubles, time, and cost, partly due to the ways the pharmaceutical industry has gotten used to interfacing data.
Fast forward to 2024 and manufacturers across industry are embracing initiatives for the Industrial Internet of Things (IIoT) and setting visions for Digitalisation and as part of ‘Industry 4.0’. But what has changed today in how we connect devices and systems together? As we are aiming for a Digital factory of the future, integration is still a major challenge but there are new ways and approaches out there that Pharma can leverage to move into the modern times of data connectivity. One of these methods is called Publish and Subscribe (PubSub) messaging and a common protocol that is growing at a fast rate to achieve it is MQTT.
What is MQTT?
Originally developed in the late 90’s by IBM, MQTT is not new, but has been gaining widespread interest and adoption due to its simplicity, reliability, and scalability. It is well suited to sending data to and from the cloud, making MQTT now a de-facto protocol for the large Cloud Service Providers (AWS, Azure etc.) to support IIoT offerings. The MQ in MQTT stood for Messaging Queue, and this hints at how the protocol works and makes it well suited for IIoT, the following are key points to understand about MQTT as a foundational aspect of your integration technology:
1. Publish-Subscribe for asynchronous communication:
- MQTT enables an asynchronous communication between devices and applications by Publishing data and Subscribing to consuming this data via a Broker. This decoupling allows for efficient data distribution without direct, point-to-point connections, enhancing scalability and flexibility.
- In the Pharma industry, heavily engineered solutions such as MES and LIMS have earnt a bad reputation for the huge numbers and complexity of interfaces. The trouble largely stems from relying on point-to-point interfaces, when a single change can result in knock-on impacts to all other connected systems, holding back the ability to easily add more data consumers and producers. Point-to-Point integration has been one of the major project and operational headaches for Pharma factories that use integrated systems and a significant hurdle to digitisation.
2.Interoperability and wide cross industry use-case adoption:
- MQTT’s open-source nature and widespread adoption has led to extensive support across hardware platforms, programming languages, and cloud services. This interoperability enables easy communication between disparate systems and devices, and is increasingly considered as standard by the large CSP’s, equipment, and software vendors.
3.IIoT specific versions of MQTT:
- MQTT has been extended specifically for use in industrial automation projects with features for auto-configuration, and to handle exception cases such as when a device is not available or goes offline. The dream of ‘plug and play’ in Pharma has never been reached despite several attempts using legacy protocols such as OPC. Whereas MQTT extensions such as Sparkplug, haves both the broad adoption at the cloud layer, and inbuilt ability to listen, auto-connect, and configure to new devices – this will be critical to the factory of the future where huge numbers of data producers are expected.
4.Ready for cloud and advanced analytics:
- As MQTT is now so prevalent and well suited for cloud applications, data that is converted to MQTT is easily prepared and accessible by analytics packages and AI/ML tools. AWS and Azure both provide MQTT as a standard protocol in and out from their powerful IIoT and data analytics services, so customers can move forward faster with initiatives to connect data sources, and utilise them in dashboard and machine learning tools.
- Using traditional approaches to integration common in the Pharma industry is now becoming a larger gap to the modern analytics world. Increasingly, industry efforts to standardise have instead created more integration silos and costly proprietary interfaces than the cloud’s move to open and easily deployable methods.
Summary:
Legacy software and interfacing approaches have created a high complexity and cost in Pharma IT projects. It has been common to see 100’s of person days spent in effort and time to setup, test, and maintain what should be a simple device.
Today, cloud native solutions like BatchLine Lite MES use MQTT to integrate to weighing scales and other measurement equipment with only a low cost IoT gateway that is quick, secure, easy to validate, and enables data to be accessible by latest cloud analytics.
