The first Programmable Logic Controller (PLC) was created in 1969 by Dick Morley and his team at Bedford Associates, meeting General Motors’ need for a more flexible and reprogrammable control system than traditional relay panels. The result was the Modicon 084, the starting point of modern industrial automation. With a modular architecture and software-based logic, the PLC replaced electromechanical circuits with digital programming, optimizing maintenance, reducing costs, and increasing the adaptability of processes.
Since then, PLCs have evolved with increased processing power, integration into industrial networks, and support for standardized languages. In this scenario, technical standards are essential to ensure interoperability, security, and efficiency. The highlight is IEC 61131-3, which defines five programming languages — LD, FBD, ST, IL (obsolete) and SFC — promoting portability between platforms. IEC 61131-2, on the other hand, deals with the electrical and environmental requirements of PLCs.
In addition to the general standards, there are those of specific application, which adapt automation to the demands of each sector. Pack ML standardizes machine states and is common in the packaging industry. The Weihenstephan Standards (WS) organize data collection in the food, beverage and pharmaceutical sectors. ISA-88 and ISA-95 are used in the automotive sector for batch control and integration with management systems. IEC 61850 is applied in energy, standardizing communication between protection and control devices.
In the field of communication, protocols such as Modbus, Profibus/Profinet, EtherNet/IP, CANopen, DeviceNet, and OPC UA are widely used to connect PLCs to sensors, actuators, HMIs, and supervisory systems, ensuring reliable and real-time data exchange.
Inside the Machine: Pack ML
Pack ML (Packaging Machine Language) is a technical standard developed by OMAC based on the ISA-TR88.00.02 standard, aimed at standardizing operational behavior and communication between machines. Its objective is to establish a common language between equipment, regardless of the manufacturer or the automation platform used.
This pattern defines a model of operational states, such as:
- Idle: Resting state, waiting for command to start.
- Starting: The machine is preparing to go into operation.
- Perform: Active state of production or normal operation.
- Stopping: Controlled stopping process of the machine.
- Aborted: Unexpected outage or critical failure.
- Resetting: The process of resetting after a failure.
- Held: Intentional pause, usually by operator intervention.
- Suspended: Automatic pause due to external conditions.
These states organize the machine lifecycle based on the PackTags data structure, which standardizes status and control variables, making it easier to integrate with supervisory systems. The modularity of Pack ML also allows the use of reusable function blocks, optimizing software development and maintenance.
A practical example is a beverage production line with a filler, labeler, sealer and packaging machine from different manufacturers. With Pack ML, they all follow the same state model and data structure, allowing plug-and-play integration, standardized interface, and unified monitoring. This reduces commissioning time, facilitates training, and improves performance analysis, such as OEE (Overall Equipment Effectiveness) calculation.
This organization format allows for a clear and automated separation between active production times (Execute), planned (Held, Stopping) and unplanned (Aborted, Suspended). This ensures that availability, performance, and quality indicators are calculated more accurately, reflecting causes of efficiency loss and facilitating corrective action.
Standardization also allows for more reliable diagnostics, comparability between lines, and automation of data collection.
Inside the Machine: Weihenstephan Standards (WS)
The Weihenstephan Standards were developed under the guidance of the Technical University of Munich (TUM), with a focus on sectors such as food and beverages. WS establishes how data should be acquired and transmitted from machines to higher-level systems.
One of the pillars of WS is the definition of more than 440 standardized data points, organized by industry-specific machine profiles. These profiles — such as WS Pack (filling and packaging) and WS Food (food) — ensure that each type of machine provides exactly the data needed for control and analysis, in a structured and consistent way. The latest versions of WS adopt the OPC UA information model, which expands interoperability, security, and scalability.
Each data point represents operation information, such as machine status, production counters, operating times, quality data, batch information, operating modes, and alarms. This data is grouped by more than 150 machine classes, with mandatory and optional points.
The Machine Status data point indicates whether the machine is in operation. This status directly feeds into performance analytics and real-time operational decisions. Standardized states in WS include:
- Producing: normal operation, valid products being manufactured.
- Setup: preparation for a new batch or product.
- Standby: machine ready, but waiting for command.
- Unplanned downtime: unexpected failures or interruptions.
- Planned shutdown: scheduled maintenance or cleaning.
- Off: out of operation, end of shift or maintenance.
- Manual Mode: operation under direct operator control.
- Maintenance Mode: technical interventions in progress.
- Alarm/Fault: Critical condition that requires immediate action.
The organization of machine states — made in standardized codes (e.g. 0 = Producing, 1 = Planned Stoppage, etc.), Boolean signals (e.g. MachineRunning = true/false) or descriptive strings (e.g. “Producing”, “Idle”, “Error”) — directly impacts the efficiency of the production line and the calculation of OEE:
From Data to Decision: Comparison between WS and PackML
Connected industries use technologies such as IoT to integrate sensors, systems, and devices via IP and TCP protocols, allowing continuous collection of operational data. This connectivity, aligned with PackML and Weihenstephan standards, ensures standardization, interoperability, and real-time control. Systems like Historian store and analyze large volumes of data, recording variables over time for accurate diagnoses.
Data science, applied on this standardized basis, enters the flow to generate insights and optimize the production process, reduce costs, and allow predictive maintenance.
Two of the most used standards in this context are WS and PackML. Although both aim to structure the behavior of machines, their approaches and levels of detail are quite different — and this is directly reflected in the way they impact a calculation such as OEE.
The comparative visualization between the standards highlights this difference. Each row represents a WS machine state, with its functional equivalent in PackML. The color-segmented bars—blue for availability, green for performance, and red for quality—show how each state contributes to the three OEE pillars.
Comparing the color tones of the graph, what stands out is the granularity of the WS. While PackML groups similar behaviors into states such as Held, Suspended, and Aborted, WS separates them into categories such as Planned Stop, Unplanned Stop, Manual Mode, and Alarm/Fault. This subdivision allows you to capture operational nuances more accurately, assign distinct impacts to each type of downtime, and improve traceability and analysis of causes of loss.
For example, in WS, Planned Shutdown and Unplanned Shutdown are distinct states, with different impacts on OEE. In PackML, both can be represented by generic states, without explicit distinction of cause. This is not a limitation, but rather a consequence of the functional focus of each pattern.
WS is oriented to the standardization of data for production analysis, with a focus on KPIs, reports and integration with supervisory systems. PackML, on the other hand, is focused on machine automation and control, and prioritizes interoperability between equipment and state logic for sequence control. Therefore, WS needs to be more detailed — it answers questions such as: Which type of downtime most affects availability? Which mode of operation is associated with the greatest loss of quality? How does setup time impact performance?
This comparison shows how different control and standardization architectures influence how data is interpreted and used for continuous improvement, operational efficiency, and data-driven decision-making.
Learn more about ST-One.