Big data and cloud for manufacturing
Industry 4.0 and Industrial Internet of Things (IIoT) discussions often focus on advanced connectivity. While connectivity is important for implementation, the entire enterprise infrastructure and system architecture also should be considered. For many years, businesses of all sizes have maintained on-site server hardware that hosts e-mail, databases, and other applications, such as enterprise resource planning (ERP) systems. In today’s hyper-connected world, even some of these basic electronic services are moving to the cloud, using Internet-based services rather than installing and maintaining servers locally. These offerings are known as software as a service (SaaS). Similarly, cloud-based services also are being adopted for IIoT and Industry 4.0 implementations.
So why use the cloud, and how is it being used? One of the more prevalent applications is when machine builder original equipment manufacturers (OEMs) and the equipment end users leverage cloud-based services to monitor machines remotely, across widely dispersed facilities. These implementations could span several enterprise networks to log this information to cloud databases, as well as collect and analyze energy consumption data, production data, or even direct sensor values, providing significant value for manufacturers. The insight gleaned from the analysis of this data also yields actionable information that populates the dashboards of decision makers so they are well-armed to intelligently chart the course of facility improvements and manage predictive maintenance efforts. Vital information and notifications, such as alerts sent when a production line is down or when machines will soon need attention, can be sent from cloud-based messaging services in the form of e-mails and push notifications to mobile devices.
Cloud-based services for industrial applications are growing at a rapid rate, and the associated technology is evolving at a parallel velocity as these innovations find use in applications with far greater scope than just the IT space. Services can be useful for industrial data collection and analysis. One cloud-based hub for IoT creates a simple, fully scalable method to connect industrial equipment and devices to that hub’s cloud services. Modern controls vendors are implementing leading IoT protocols, including MQTT [a messaging transport protocol] and advanced message queuing protocol (AMQP), to further streamline cloud-services integration. The barriers to entry when initiating smart factory development are far lower when these next generation cloud concepts are applied in industrial applications.
Data, where needed
With more connected devices and more data being collected, manufacturers need advanced analytics to help convert the data from raw figures into actionable information as well as an operations plan. Basic machine analysis includes threshold monitoring of digital and analog values and operational timing analysis: total, minimum, maximum, and average, program state analysis, and energy usage calculations all fall into the set of analyzed data. This data can be stored in standardized formats with data compression, either locally in the controller, in a cloud-based solution, or on a server within the enterprise network or in a public cloud as business needs dictate.
Further benefits arise from data analysis, and one example can be found with predictive maintenance insights. Logging data from operating hour counters, frequency analysis, or root-mean-square (RMS) calculations, for example, enables implementation of high-performance condition monitoring. The system facilitates limit value monitoring for different process data. Pattern recognition for detecting irregularities and repetitions in the recorded data further improves process-sequence reliability.
Analytics and OEE
Analytics software offers numerous opportunities to modernize machinery. Status analysis provides the information required for optimizing the machine or system for overall equipment effectiveness (OEE). Detailed knowledge of processes simplifies drive configuration and may help reduce the connected load of machines based on comprehensive measurement readings.
Analytics tools, such as loggers for recording data, software-based scopes, function blocks for features such as RMS calculations, or simulation software platforms serve an increasing demand for maximum data transparency. In this case, analytics can provide comprehensive production documentation for each product, down to batch-size-one production.
In a real-world use case, a machine OEM configured all machine controllers with the ability to send operational data from the machine to the cloud, over one Internet connection. This data included alarms, runtime and downtime, average cycle time, controller CPU temperature, and, optionally, running product size.
Obtained information drives innovation into the next generation of machine design. Knowing the most common alarm across a widely installed base of the machine type can provide the OEM with priceless feedback on machine performance that may be unavailable via traditional methods. This allows for better allocation of engineering and design resources, as the company knows more precisely where to make improvements on future machines to offer more competitive solutions.
Better asset utilization
Additionally, having access to data, such as running product size, enables analysis of how the machines are mostly being used, and it can aid in optimizing machine operation and next-generation machine design. For example, take a machine that will run products from 5 mm to 500 mm in size. If 98% of the time the majority of the installed base only runs products from 120 mm to 220 mm sizes, that insight helps the OEM to offer a new machine model optimized for size and cost that focuses on a maximum product size of 220 mm.
Machine and equipment suppliers are bringing additional benefits to end users through predictive maintenance and condition-monitoring features as well. Collecting and managing data from machines on a subscription-based model ensures that the end user’s machines run as optimally as possible and that the appropriate people are notified when concerning anomalies are seen during operation.
Equipment suppliers are ideally suited for this brand of analytics, as they are experts in their particular equipment. This also boosts intellectual property and value proposition to remain ahead of competition.
4 ways to launch an IIoT project
For implementation, where should someone begin?
The place to start for IIoT projects is to identify the business challenges and real-world use cases. The needs are typically rooted in quality, production throughput intelligence, predictive maintenance, and overall equipment health. For example, a manufacturer may:
1. Seek to increase machine uptime, enable production managers to get real-time notifications on mobile phones when production is hit with expected complications or when product quality from a line decreases beyond a given threshold. Defining these parameters up front helps give the project definition and scope.
2. Enable the collection and streaming of the data needed to help make analytics and notifications possible. This requires identifying what database type and software analytics tools should be deployed for the task and choosing a communications protocol compatible with those tools, as well as with the machine controllers in the facility. Many emerging and powerful analytics tools for manufacturing support IoT protocols, though many existing industrial controllers still do not support them.
3. Take advantage of convergence of automation technology (AT) and information technology (IT) happening now. This is exactly where PC-based controllers help by capitalizing on this convergence and by allowing fast and efficient implementation of new programming standards and protocols, regardless of whether they originated in the automation and controls realm or in the IT world.
4. Use software and hardware that facilitate integration. Some industrial suppliers are already implementing protocols, such as MQTT and AMQP, into the control platform with access directly from programmable automation controller (PLC) languages. PC-based controllers also can be used as small, flexible gateways that can connect to traditional PLCs using the legacy fieldbus protocol and can translate that data into IoT protocols, then send that data to local or cloud-based servers. This eliminates "islands of automation" in the emerging smart factory, even for plants that can’t yet completely do away with the more limited, traditional hardware PLCs.
Yes, it’s really happening. IIoT and Industry 4.0 strategies are taking a more physical presence as companies begin to capitalize on tangible benefits. Machine optimization and continuous improvement efforts remain key elements in growing a global company, and PC-based control and next-generation analytics are helping the most competitive manufacturers harness big data to reach the ultimate destination of the smart enterprise.
Software helps with cloud-based connections, IIoT implementations
Various software packages help with Industrial Internet of Things (IIoT) implementations.
- E-mail, databases, and other applications, such as enterprise resource planning (ERP) systems can span several enterprise networks and log information into cloud databases, such as Microsoft Azure SQL. Applications can collect and analyze energy consumption data, production data, or even direct sensor values, providing significant value for manufacturers.
- Microsoft Azure offers several services useful for industrial data collection and analysis. The Azure IoT hub creates a simple, fully scalable method to connect industrial equipment and devices to the Azure cloud.
- Controls vendors, such as Beckhoff Automation, are implementing leading IoT protocols, including MQTT and AMQP, to streamline integration of cloud services.
- In addition to error-analysis support, Beckhoff Automation TwinCAT Analytics software offers other opportunities to modernize machinery.
- Analytics tools, such as loggers for recording data, software-based scopes, function blocks for features such as RMS calculations, or platforms such as Matlab/Simulink from The MathWorks, serve an increasing demand for maximum data transparency.
– Daymon Thompson is automation product specialist, Beckhoff Automation. Edited by Mark T. Hoske, content manager, Control Engineering, firstname.lastname@example.org.
- Cloud-based software as a service can host applications to optimize industrial machines.
- Industrial Internet of Things (IIoT) projects are easier with integration tools.
- Optimization and continuous improvement advance with analytic tools.
IIoT and Industry 4.0 initiatives aim to improve manufacturing productivity; are you implementing them?
Beckhoff Automation’s Website has an Industry 4.0 section. Search "Industry 4.0" at www.beckhoff.com.
– See additional stories by the author linked below.