Smart Manufacturing strives for higher levels of connectivity in the enterprise with information flowing in near real-time between production and business systems to achieve highly orchestrated physical and digital processes within plants, factories and across the entire value chain.
In this article we discuss a bit more about the infrastructure bones you want to futureproof your Smart Manufacturing strategy and provide security, interoperability, and scalability as you evolve your systems. This Smart Manufacturing infrastructure includes the integration of the operational technology (OT) running machines and automation in the plant with the information technology (IT) running the business and providing tools like simulation, advanced analytics, and machine learning platforms to enable real-time semi-autonomous control and optimization of production processes.
Smart manufacturing requires data from many different things in the plant including machines, tablets, wearables, and sensors. Different things and systems are designed to exchange data in different ways for different purposes.
The systems controlling machines need to act on data within fractions of a second, however people and systems running the business usually act on data between days or shifts. The different technologies and systems needed to run these very different processes lead to natural layers within the Smart Manufacturing system landscape dedicated to (i) connected things, (ii) connected systems, and (iii) connected teams.
Smart Manufacturing requires a flexible systems architecture to automate activities, pass data seamlessly, and enable new levels of optimization, predictive and prescriptive analysis in manufacturing operations management and enterprise processes—a systems architecture that is interoperable, scalable, and futureproofed for easy upgrades as technology continues to evolve.
Connected Things
Smart Manufacturing must provide broad, secure connectivity among devices, processes, people, and businesses in the ecosystem, securing data integrity, protecting intellectual property, shielding against cyberattacks, and maintaining business continuity with minimal impact to performance of the overall network of networks. Security includes identity verification schemes for every “thing” in the ecosystem, implementing schemes for access control, data traffic monitoring, software patches, fault-tolerance, high availability, anomaly detection, issue containment, and data recovery.
Data integration for production automation traditionally involves proprietary application interfaces from specific technology vendors and leaves manufacturers with hundreds of point-to-point connections that only a specialized group of people understand. To avoid having data trapped in proprietary data silos, organizations can organize data in a distributed architecture leveraging technologies like data lakes, data platforms and open standards-based information models.
Smart Manufacturing systems should implement interoperable solutions to empower a connected ecosystem of devices, systems, and people communicating in a natural yet structured manner. The manufacturer must be able to engage in B2B (business-to-business) data exchanges with the supply chain management systems of ecosystem partners with data preferably collected directly from production systems and transmitted via M2M (machine-to-machine) and A2A (application-to-application) integration APIs (application programming interfaces) that increasingly are more open and standards-based to enable application portability and multi-vendor hardware and software plug-and-play solutions.
Information models are used to contextualize and organize the data from production processes along with the relationship to ingredients, resources, and production orders. Information models enable data to be shared for multiple purposes among multiple stakeholders in the enterprise.
For example, captured data can represent a change in temperature, a completed batch, a drop in production line efficiency, or a customer complaint under a specific level of the information hierarchy. When the data captured from processes is contextualized and communicated using information models, the applications and users receiving the data can understand the context of the process it came from for a meaningful analysis.
OPC UA (open platform communications unified architecture) is a popular standard for information modeling with a growing library (OPC Cloud Library) of reusable information models for specific types of machines and industrial processes. For example, a process engineer can download an OPC UA information model for interoperability among food processing equipment instead of creating their model from scratch. Not only does the engineer benefit from the prior work of collaborators on the standard, but they also benefit from applications that have been designed to work with the standard.
When applications consume data via information models and open APIs from a shared data platform, data producers and consumers are not directly connected. Instead, they communicate through a mid-tier data platform that enables the organization to easily replace applications without breaking all the data connections. Organizations can continue to add more AI applications that monitor data, detect patterns, and trigger actions and alerts based on predicted outcomes.
Connected Systems
Smart Manufacturing systems must be scalable across all functions, facilities and the entire value chain with cost growing linearly—instead of exponentially—as load and complexities increase. Cloud computing, virtualization, and containerization techniques allow performance to be maintained by distributing the workload and scaling of computing power as the needs of the organization evolve. Edge computing solutions make it possible to place latency-sensitive real-time control applications on-premises close to the machines while other production applications, like metrics dashboards, take advantage of cloud services for scalability and ecosystem connectivity.
Modular applications (a.k.a. apps) are easier to maintain and replace than larger systems with a broad functional footprint. The combination of interoperable and modular solutions makes the architecture more future proof allowing systems, components, and resources to be added, modified, replaced, or removed with ease to accommodate the changing demands.
For example, a modular app that tracks job starts and stops in production might send updates to both a manufacturing execution system (MES) for job status and to an enterprise resource planning (ERP) system for tracking employee labor time sheets. To increase interoperability, smart manufacturers will typically standardize on the use of APIs to transmit data among these internal systems.
Low-code development platforms have evolved to make it easier for organizations to create the modular apps they need with their own OT-IT personnel. Low-code platforms can empower users to build custom solutions that improve their operations while the IT department maintains architectural consistence and governance. Operations and IT teams can develop a better balance and collaboration with these types of solutions. However, organizations must still evaluate build versus buy decisions, and consider the long-term maintenance of internally developed custom solutions.
The integration of OT and IT is an essential step in the journey towards creating a fully connected, dynamic and flexible Smart Manufacturing systems architecture. It’s not a small step to take, but without it, the ability of a manufacturer to transact and participate effectively in a highly connected manufacturing ecosystem will be limited.
Integrated IT and OT systems enable more streamlined processes across production and business functions. For example, an IIoT sensor can collect operational data on a machine at the factory and send it over a wireless network to an IT application that performs predictive maintenance analysis. That application can trigger a maintenance order in the maintenance system to dispatch a mechanic and perform maintenance on the machine to avoid the potential of a longer downtime due to an unplanned machine malfunction.
The reality for many industrial organizations is that they have implemented multiple generations of equipment and systems that should all be connected. The good news is that organizations can gradually transition in their journey to a fully integrated OT-IT systems landscape. There are methodologies and technologies available that allow a smooth gradual transition. A rip-and-replace approach is usually not the recommended approach.
Companies that master the OT-IT convergence and implement Smart Manufacturing techniques ahead of their peers will have a competitive advantage in highly digitally connected supply chain including quicker visibility into issues that would affect the delivery of materials, parts and products, and quicker resolutions of those issues.
Connected Teams
To reach the goal of Smart Manufacturing, the organization must not only implement technology to improve the business it must make sure people are ready for new technology-enabled processes and organizational change.
Even in the most automated plants humans must interact with machines to set up, monitor and control the manufacturing processes, ensuring that they are working properly and producing a quality product. People interact with equipment through a human-machine interface (HMI) and interact with applications through a user interface (UI).
An HMI delivers information from machines to operators, allowing them to control, monitor, record, and diagnose machines. HMIs can be integrated into a single machine or a collection of complex devices that must all work together under one control system architecture commonly referred to as supervisory control and data acquisition (SCADA). An HMI can monitor locally or remotely real-time data from a process like water level, the pH, the water pump, the level of dissolved solids or a certain toxic chemical. The water pump can be turned on or off based on tank levels using the HMI. In addition, the HMI can display alerts if the pH is below a certain level, and this can be adjusted using the touchscreen display.
A UI is the way we work with apps designed to perform tasks like managing or collecting data from assembly operations. Many modern-day industrial UIs for the smart factory are multimedia rich. They allow users to receive integrated SMS alerts about the status of machines, email alerts, and view work instructions with integrated videos of the production processes. UIs can be graphical dashboards at the desktop or phone with plant related metrics or be more specialized on a wearable device like a scanner mounted on a glove.
Whether it is an HMI or UI, human and machine interaction is a key part of a Smart Manufacturing system. These interfaces empower end-users to make better decisions with access to accurate real-time data that with visualization optimized to highlight conditions important to running each specific production process. As technology continues to evolve to assist the manufacturing worker, we can expect more voice activated interfaces and AI driven virtual assistance features will be incorporated into the UI for the factory worker.
Many organizations are still working with manual workflows defined in standard operating procedures (SOPs) that document how employees fill out paper forms and route them to multiple departments for disposition and approvals. Some have started implementing semi-automated workflows that integrate applications and tasks across departments to automate repetitive tasks and routine decisions.
Smart Manufacturing aims to streamline business processes and workflows across departments that support production operations, so they are in synch and optimized. Workflow solutions can be used to integrate modular applications, facilitate process orchestration, and integrate the tasks needed to capture data, contextualize, analyze, and trigger needed actions. For example, a supply chain transaction may need all information related to a certain customer, which could be in many different segments of the enterprise. Workflow software can collect all the information and respond automatically, which saves employees from multiple tedious manual steps. Organizational change may also be required to achieve improved streamlined business processes and get products designed, outsourced, built, tested, packaged and delivered to the customer in a consistent manner.
This might all sound like a lot of change for your organization but remember that Smart Manufacturing can be done gradually in multiple stages. For example, a smart manufacturer might first install real-time capabilities for production monitoring. After this, they could integrate smart measurement devices for quality tracking, and later, systems to track raw materials as they are used in production to avoid material shortages.
However, it is important to establish a shared data architecture early in the Smart Manufacturing journey along with guidelines for connectivity, information models, security, data exchange, and scalability. These shared practices will ensure that each technology adoption project is contributing to the overall Smart Manufacturing vision for the company.