Breaking down the IT and OT barrier is maximizing productivity
Digital technologies are transforming the management of manufacturing, distribution and utility services. Combining operational information from sensors with business data from the supply chain and customers makes it possible to react to changes quickly.
Organizations transitioning to Industry 4.0 are seeing benefits in several areas, like servicing demand more efficiently. The benefits extend to improving product quality and sustainability. If organizations bring together the deep experience and skills of the groups that keep the business running, they can tap all the potentially available benefits.
The teams handling operational technology (OT) and information technology (IT) need to work together more closely. Each brings deep knowledge in its area. The OT team’s core skills will focus on maintaining the shop floor systems, from basic actuators and sensors through machine tools to robots. And they will often have a deep involvement with the design and creation of these systems.
The IT team’s systems managing the wider enterprise deliver a higher-level view. Billing, order management and scheduling as well as the various support systems for human resources, finance, sales and customer service can often seem to have a distant relationship to OT. But information from many of these IT systems will inform the best way to maintain and schedule OT actions.
Digitalization empowers automation
IT integration will pull data from disparate industrial sensors. (Source: Avnet - https://my.avnet.com/abacus/resources/article/evolving-component-technologies-will-make-industry-4-0-viable/)
OT and IT data integration improves efficiency and profitability for just-in-time manufacturing
Consider a business that operates a just-in-time supply chain. When the manufacturer receives orders from customers, it will need to adjust its requests from component suppliers and make changes to production scheduling on the shop floor. Those changes make sure products are on the way to the customer as quickly as possible. It is not just a matter of speed but of providing better information to customers.
The manufacturer or distributor can send accurate delivery estimates using data collected from workers’ hand scanners and picking robots in the warehouse. The updates continue after the product leaves the manufacturing shop floor or warehouse. GPS readings from delivery trucks will also form part of the data collection process, handled by systems that convert the real-time OT information into a form the IT systems can use. Companies that have taken full advantage of this data-sharing have become leaders in their field, leaving behind competitors struggling with the overhead of traditional systems.
As well as improved customer relations, a closer integration between OT and IT can yield other improvements that deliver improved profitability. The cloud computing power available to IT teams enables tools that improve throughput and reduce downtime. According to IDC's 2022 Worldwide IT/OT Convergence Survey, the average cost of unplanned downtime in manufacturing operations costs to companies was almost US$220,000 per hour.
IT's involvement here is important. It may implement AI models that run in the cloud, trained on data from the OT domain. AI models can identify signs of excessive wear in machine tools and other equipment. The output from these models helps schedule maintenance during quiet periods and shift changes. That improved schedule minimizes the impact of any repairs that will cause equipment to be taken out of service. Inspection systems added to the production line can also watch for changes in product quality that may signal issues with upstream machinery. This adds an important feedback loop to the preventive-maintenance systems by helping to update the AI models that assess how well each machine is running.
Protecting industrial automation from cyberattacks
Industry 4.0 requires an increased focus on communications security. (Source: Avnet - https://my.avnet.com/silica/resources/article/industrial-iot-security-why-and-how/)
Digital twins maximize production management in Industry 4.0
The digital-twin concept fully embraces IT systems to monitor the long-term conditions of operations. Organizations are creating complete interactive models of their plants, facilities and pipeline networks.
Some manufacturers have already used the technology to plan the best shop floor layout for their equipment before building and equipping the factory. The ways technology helps include:
- 3D graphical modeling to evaluate different configurations of picking robots and aisles, minimizing holdups and speeding up the loading/unloading of products.
- Using models of their networks to gauge where to devote resources to improve operational efficiency.
- Actioning data from inspection systems and the machine tools themselves to provide information that is needed to ensure the model within the digital twin remains consistent with reality.
Using technology in this way will support better decision-making. Management can be sure that a change indicating better productivity in the model will be reflected in the physical domain.
Production management, digital twins and preventive-maintenance systems do not just rely on bringing data from the OT world into the IT domain. They also benefit from the streamlined flow of data from IT databases into the OT domain. Though easy to state, this is a transition that is often difficult to achieve if not addressed strategically. Though better network connections represent a major leap to enabling Industry 4.0, it is important to address differences in how the two halves of the organization look at problems.
An underlying difference lies in the handling of time-sensitive information. Real-time considerations dominate the OT world. Decisions by a machine tool, actuator or robot need to be made relative to deadlines that are often measured in milliseconds. IT systems and many of the AI technologies that could augment these systems do not have as rigid a concept of time.
Any system that combines elements from these two worlds must consider the impact of time and deadlines. The system needs to be prepared to reprioritize data flows or annul them if they do not arrive in time to be actioned effectively. A similar understanding of the differences must be part of the planning and implementation process used by the joint OT and IT teams.
Optimizing data management for industrial networks
There are other practical challenges. Shop floor networks and fieldbuses are optimized for time-sensitive and often small packets of data. The fieldbuses will also have limited coverage. A programmable logic controller (PLC) may just connect to a few machine tools within a cell with little to no interconnection with other cells.
Where there is a more extensive network, gateway modules may pass high-level commands and status updates along a backbone that connects cells. But this might not be a fully standard form of Ethernet. As a result, it may not be possible without changes to the network or the gateways to retrieve useful data for use by a digital-twin or AI models.
Security presents a second issue. Many industrial networks rely on air gap. A physical break provides isolation from remote attackers. As soon as the backbone network on the shop floor is connected to the internet, sending data to the IT organization’s cloud, it becomes vulnerable.
The data exchanged by machine tools and PLCs will often be pared down to its bare essential to avoid consuming too much precious capacity on fieldbuses. When pulled into the cloud from the shop floor network, IT systems will find this data difficult to interpret. The proprietary data formats that traditional PLCs typically employ complicate the situation. The metadata or formatting used by one PLC to present the data they produce may also be inconsistent with the systems used in other production cells. As a result, if the data is available to IT systems, it may prove very difficult to employ directly if not converted to a standard form with metadata that is recognized across the entire enterprise.
Additional systems to deal with security and data-translation tasks may be necessary. Firewalls between the OT backbone and the IT network can block all traffic from the internet other than packets that are verified to be authentic. Some IT teams use zero-trust strategies, where each message is verified individually for rights and provenance. These checks help ensure that hacking attempts do not pass through the firewall.
Solutions for retrieving shop floor information, adding data conversion and contextualization services, are now available. Avnet’s IoTConnect® provides services that help capture and transmit actionable data from sensors and IoT devices to IT systems in the cloud.
Other OT systems added to the industrial network may harness AI and edge computing to provide real-time inspection services for quality control. One way in which OT teams are integrating these systems into their networks is to move to a new generation of flexible PLCs with more advanced services and support features such as those outlined in the Open Platform Communications (OPC) Unified Architecture (UA). This set of standards implements a publish-subscribe model for industrial data and supports data encryption combined with authentication and role-based access control to prevent data from being misused.
Fragmentation into silos can prevent organizations from capitalizing on their strengths in OT and IT. By focusing on collaboration and putting in place systems that allow these two areas to work together effectively, manufacturers and service providers can implement new techniques such as digital-twin modeling and use them to leapfrog their competition.
More industrial-related articles from Avnet:
