Why an MES and Industrial Data Platform Strategy is Critical
Manufacturing execution systems technology has been around for decades and has been experiencing a resurgence in interest as industry looks to capture and analyze detailed production data for insights into operational capabilities and improved business decision-making. Accompanying this interest around MES technology is an increasing focus on applying industrial data platforms to more effectively manage all the operations data being captured.
To get a better sense of these trends around MES and IIoT data platforms and where they interact, Automation World (AW) connected with Francisco Almada Lobo (FAL), CEO of Critical Manufacturing, a supplier of MES and smart factory software, including industrial data platforms, augmented reality, digital twins and advanced planning and scheduling for advanced industries such as semiconductor, electronics, medical devices and industrial equipment.
You can hear the entire podcast at https://www.automationworld.com/55289387. Following are a few highlights from the discussion.
AW: Can you describe how MES and industrial data platforms complement each other in terms of connectivity and access to real-time operations data.
FAL: What we're seeing now and are advocating for is the emergence of a three-part foundation for Industry 4.0 success. These three parts are: MES, equipment connectivity and the data platform. Each piece has a distinct but interconnected role: The MES manages the execution, equipment integration enables the real time communication between the MES and machines, and the data platform ingests high volumes of data from equipment and systems that then contextualizes and stores the data into a unified model that allows for analysis.
But the point is that when the three of these are working together, we can create real value. That value comes from knowing that equipment and sensor level data is not only captured but is being contextualized with MES information stored in a common data model. This is what enables advanced analytics, traceability and AI-driven decisions at scale and closes the loop, as the actions that result from these decisions will reach back into the equipment itself.
This is the shift that we've been seeing for some time and now it's clear that this is the path forward in terms of Industry 4.0.
AW: Considering the vast amounts of data that are generated by connected devices in a manufacturing environment, how do MES and industrial data platforms handle all of this data effectively?
FAL: The MES handles real-time instruction and actionable data for things like tracking, quality checks and operator inputs. It essentially focuses on immediate control and traceability and only stores the data needed to ensure the execution of those tasks.
The data platform is where we ingest high-volume, high frequency data from machines and systems. Then we combine this data with MES data to provide the level of context needed for analysis and decision-making.
The data platform is where we ingest high-volume, high frequency data from machines and systems. Then we combine this data with MES data to provide the level of context needed for analysis and decision-making.
Integration is the key here, because together the MES and the data platform divide the workload. The MES filters and records the actionable data for execution and the data platform captures the high-volume, high-velocity data for analysis and insight.
It's these two technologies working in tandem that ensures no valuable data is lost and each system can do what it does best — ensuring control of the shop floor and enabling continuous improvement through analytics.
AW: A big issue with managing operations data coming from multiple devices and systems involves the need to contextualize all the data for analysis. Can you explain how Critical Manufacturing uses AI for this?
FAL: Data contextualization is critical for analysis, but we don't use AI to perform the contextualization. Instead, we ensure that the data is automatically contextualized starting at the source by linking sensor and equipment data with key operational information. For example, if you have a temperature reading from machine, it can be enriched with additional information, such as the type of machine it came from, when the machine last operated, when it was last maintained, the program it is running and the kind of material being process on it.
Given this context, the raw data can then be transformed into really valuable information that can be used effectively for monitoring, analysis or decision making. With AI added you can analyze the data to detect patterns and perform regressions across machines.
By using this layered approach of applying context first, then the intelligence ensures that AI is applied to data that actually makes sense, so that the resulting decisions can have as much impact as possible on the shop floor.
This is where we train purpose-built machine learning models for specific use cases. Once trained, the models are deployed within the platform so that they can run automatically in production, analyzing new data in real time and delivering actionable predictions directly to operations.
By using this layered approach of applying context first, then the intelligence ensures that AI is applied to data that actually makes sense, so that the resulting decisions can have as much impact as possible on the shop floor.
AW: So how do MES and industrial data platforms — working in tandem and then leveraging AI — improve production and operational reliability and minimize downtime?
FAL: When MES and data platforms work together and are enhanced by AI, they form a powerful foundation for improving your personal reliability and minimizing downtime.
As an example, the MES ensures consistent execution by managing the production orders, enforcing processes, and handling real time control and traceability. The data platform continuously ingests these high frequency data from machine sensors and systems and enriches it with MES context.
The AI and machine learning models trained on this rich data set can predict potential failures before they occur. For example, the models can identify process deviations or root causes of downtime to optimize schedules and settings to avoid excess stress on equipment.
But it goes further than this. These models can also alert operators or technicians before a problem is visible. They can suggest parameter settings to trigger maintenance activities or automatically trigger responses like stopping a machine, re-routing production or executing a specific MES transaction.
This closes the loop from insight to action — sometimes even autonomously — and the result is a shift from a reactive to proactive and even autonomous reliability management to reduce plant downtime, extend asset life and improve production stability.
In addition to identifying process deviations or root causes of downtime, the AI and machine learning models can suggest parameter settings to trigger maintenance activities or automatically trigger responses like stopping a machine, re-routing production or executing a specific MES transaction.
AW: What advice do you have for manufacturers looking to integrate legacy equipment and systems into an MES and industrial data platform framework?
FAL: Integrating legacy equipment and systems is one of the most common challenges to modernizing manufacturing environments. But today it's highly achievable with the right connectivity strategy.
At Critical Manufacturing, we address this through two different approaches: This first of these is protocol level integration where we use our equipment integration layer called Connect IoT . It allows you to integrate data from different equipment using specific communication protocols. We support a wide range of protocols. This makes it possible to develop custom drivers to interface with equipment that uses proprietary or outdated protocols.
The other approach is through OPC server integration. Many legacy systems cannot connect directly to modern platforms, but they can expose their data through an OPC interface. This makes the data accessible to both the MES and an industrial data platform in a standardized way.
Another thing manufacturers should understand is that, even if their equipment does not have any IoT sensors, it’s possible today to retrofit the equipment with relatively inexpensive sensors to capture data, such as vibration or temperature or piece counting. This data can then be collected via Connect IoT using one of our supported protocols to make older machines part of a digital ecosystem.
Essentially, if a machine can produce data, then we can integrate it.
Leaders relevant to this article: