Modern Factory Integration

Christopher Murphy, VP of Software and Controls Engineering, Jervis B. Webb Company Innovation
Christopher Murphy, VP of Software and Controls Engineering, Jervis B. Webb Company Innovation

Christopher Murphy, VP of Software and Controls Engineering, Jervis B. Webb Company Innovation

Modern control systems strive to achieve more complex goals than ever. Production systems aim for higher levels of production diversity, product quality, traceability, and visibility. Producers want to produce customized product on a single assembly line. Each produced unit has a different set of features. The target is build-to-order for instant customer gratification.

The tooling, work cells, logistics systems, and material handling systems all produce data that the manufacturer is using to control the quality of the product. In addition, the data supports the processes that remediate quality problems expeditiously. For example, a tooling sensor may detect an error, and, based on the error, the product can quickly be routed to the appropriate rework cell.

 "The modern factory needs see all the measurable parameters of quantity, quality, efficiency, and rework" 

Traceability supports regulatory compliance, post production support, and product documentation. A complete record of the product production process, down to fastener torque, is accumulated. The data feeds back to design refinement, manufacturing process refinement, and augments warranty support.

Visibility into the manufacturing process promotes production efficiency. The modern factory needs to see all the measurable parameters of quantity, quality, efficiency, and rework. Workers need to see the factors that are affecting them now and what factors, such as parts shortage, will be affecting them in the future. When a line stoppage happens, the plant needs instant visibility of the issue for rapid response. Modern SCADA systems are a typical part of the integration to support visibility.

The technologies that support the modern production process include intelligent sensors, programmable controllers (PLCs), field bus networks, data networks, SCADA systems, ANDON systems, logistics systems, tooling, safety systems, and material handling systems. These systems need to be coordinated with the production planning system to produce the choreography of a well running system.

Sensor evolution is approaching an Internet of Things level of implementation. Field bus architectures make extended data available to the higher-level consumers of data. Integrators are moving away from proprietary network technologies to Ethernet for cost and simplicity of implementation.

A complex integration project requires rigorous planning for success. In a common model, the companies MRP system evaluates the materials and orders to produce a production schedule. The schedule drives the flow of materials through the system.

The integrator maps the material flows through the production process, from material order to final shipping. She maps the data flow on top of the material flow, ensuring that data record for each material flow is built and recorded.

In parallel, the monitoring information associated with the production equipment drives the SCADA for process visibility. Cycle counts, run time, cycle time, energy usage, trending, and other parameters of the production equipment drive the maintenance system. For example, after several hours of run time, the maintenance system produces the work order to perform preventive maintenance on the equipment. When the work order is executed and logged, the run time resets. This is like the change oil light on your car, but expanded to include all the process equipment.

At the process equipment level, controllers are dominated by PLCs. PLCs offer the plant support staff a well understood platform that they can maintain and in many cases implement improvements. The complexities of modern integrated systems challenge the capabilities and performance of PLCs in some applications. PLCs were originally designed as I/O logic/state engines with an easy to understand programming language.

With data now a dominant force in manufacturing, PLC manufactures have increased memory, added data instructions, and improved communications capability. This has led to a compromise. PLC manufacturers have tried to maintain the machine friendly features of the programming tools. These programming tools often do not support data handling in intuitive ways. Using these new capabilities requires more digital skills. A competent PLC designer in this world is more akin to a computer systems programmer.

Equipment manufacturers are challenged in constructing controllers for integration in plant systems. Plant support personnel often specify PLC based controllers. From a capabilities, development, support, and intellectual property standpoint, the integrator often wants to use a computer based solution. In many cases, the controller may be a data only device, with no traditional machine I/O. A traffic director in an Automatic Guided Vehicle system is typical of such an application. Such a controller would arguably be better implemented using information technology tools on a computer platform.

For vendors with this challenge, there are some guiding principles when implementing a computer based controller. Most industrial integration projects include a SCADA. The software application should implement a SCADA interface and expose the configuration and status of the software application to the SCADA. This allows the SCADA to be a unified view and console for the entire integrated system. This gives the operator confidence that he can see and control the entire system. This avoids the sense that the computer is a black box.

System testing is critical to successful integration. Most systems involve physical equipment that cannot be practically built prior to final commissioning. Practical testing involves a computer emulation of the physical system that the software can test against. Emulation testing is a step that requires planning, with schedule time allowed. Emulation testing significantly reduces the commissioning risk and ensures a predictable and smooth startup. Commercial emulation packages are readily available and worth the effort.

Plan your system flow and your data flow. Provide as much visibility into the operation of the pieces integrated. Test the entire integrated system through emulation. The hallmarks of a well implemented system.

Read Also

Industrial IoT is Different

Industrial IoT is Different

Wes Mukai, CTO, GE Transportation Digital Solutions
Industrial Data Analytics: An Edge to Cloud Strategy

Industrial Data Analytics: An Edge to Cloud Strategy

Jay Allardyce, COO, GE [NYSE:GE]
Digital Disruptions: SMART Manufacturing Reimagined

Digital Disruptions: SMART Manufacturing Reimagined

Sanjay Choubey, Global Program Delivery Executive, Briggs & Stratton