Tips for operational efficiency

By Control Engineering Staff October 6, 2005

I gave tips for improving operational efficiency, based on Control Engineering coverage and my observations, in a presentation at a local users’ conference recently. Lively discussion ensued, including suggestion that I include a summary of the advice in print. My October 2005 “Think Again” column will appear at /archive , and here are more tips on improving operational efficiency that wouldn’t fit in the print space.

1. Sensing. Will sensor information remain local or should results travel beyond local I/O devices? When designing, don’t leave sensors as the last item. Smart and very small sensors offer certain specialized benefits for particular applications. Sometimes a less-expensive sensor might have solved the problem, if adequate mounting space was allocated. Sensor mounting can allow applications that were previously impractical. Some automotive industry applications mount cameras above the robotics, allowing time for image processing, as the robot arm cycled through getting, placing, and returning for more parts. Mounting the camera on the robot arm would have slowed throughput. Questions to ask include: Does feedback need to be shared among multiple discrete inspection, rejection, and other stations, or will one controller coordinate inspection results with ancillary tasks? How will inspection results be used and communicated? If I/O devices are used for communication, what types and locations are required to function with other controls? Also consider lighting, power, and data transfer.

2. Networks. Networks save communication translations and machine-to-machine (M2M) data flow, and connects equipment to other parts of the enterprise and beyond. More products are connected to networks. RFID can help carry valuable information through products’ lifecycles. Ethernet is growing rapidly, offers power with signal, can be made deterministic, and is virtually future-proof. Sensors are freed by better wireless technologies. Logic can be anywhere: Embedded on a board, on a chip, in a system or stand-alone, centralized or decentralized, within a unit, cell, or networked: plant floor, plant-wide, enterprise-wide, supply chain, and Internet. Actuation is smaller, more powerful, and can be more decentralized. I/O modules can include the whole control loop in close proximity without control room or enclosure.

3. Logic/analysis: Logic choices have expanded. Get smarter: specialized tools require experts to install, and they can be difficult to integrate and upgrade. Plants often install systems, but do not use them effectively. Two examples: 90% of plants do not use information from their predictive diagnostics tools (Emerson Process Management); companies often install maintenance software without a clear idea of what they are trying to achieve (ABB).

4. Actuation: Upgrades should seek opportunities for jumping a generation beyond technologies under consideration. Rethink technology selection and design locations; design methods and tools things have changed; simulations speed design.

5. Processes: Find ways to maximize the value of plant-floor knowledge; increasingly, systems include or connect to a real-time process historian. Rather than designing something, consider purchasing power-related (and other) systems, to save time and effort. Modular design can allow faster upgrades, and shorter development times. Sub-assemblies speed the process with pre-testing. Power supplies, buses, management systems, and intelligent components all can report into the HMI for greater integration and efficiencies. Isolation, monitoring, and shielding can minimize power-line and EMF disturbances. Think about the system. Consider a machine, line, or process area in context of surrounding production flow to improve efficiencies. Seeing the entire production process helps focus resources in areas needing attention the most. Integrating information from design to machine tools through product creation and distribution yields benefits beyond the machine; savings of 50% and more can result. Making safety and security part of standard ongoing processes can add to the bottom line, rather than be considered just a cost.

6. Standards: Standards development seems to happen faster, making them more useful. (Perhaps we learned from the 15-year single unified global fieldbus standardization effort, which, in hindsight, doesn’t seem global, single, or unified.) When coding, think modularly, and find ways to carry the intellectual knowledge from the design into application code, and beyond, throughout the product lifecycle.

7. Safety devices: More safety devices are being purchased in North America and Europe, according to data from ODVA and Intex Management Services. In percentage points, the most desired safety technologies are safety networks, safety PLCs, and redundant safety PLCs.

8. Migration paths. Give greater attention to migration paths and on automated upgrades; technologies have a way of hanging on. Relays, DCS, PLCs, PCs, and process automation controllers are all in the mix in many facilities. Architectures and standards can help buffer users from the pace of Microsoft migration and plug into capability of continuous upgrades, rather than continue unimproved with the old “set it and forget it” mentality. We’re entering the age of ongoing upgrades that enhance capabilities more regularly.

9. Training, research, and references: Ensure organizational knowledge is passed on, and that industry best practices are applied.

10. Partners: Users want vendors to help make migration paths clear; preserve and maximize the value of existing assets; speed set-up; make things easy to use; deliver the most for the money; and save jobs. Partners vary widely: Single vendor, multiple vendors; custom engineering; commercial off the shelf (COTS); in-house engineering; system integrators (stand-alone and affiliated); and‘Coopetition’ (interconnecting supply chains make today’s competitors tomorrow’s cooperative partners).

More on these topics are available at www.controleng.com and the Control Engineering Resource Center (a paid subscription micro site) at /

—Mark T. Hoske, Control Engineering, MHoske@cfemedia.com