Industry Articles (50 articles)
There’s an old adage in TQM circles: “You can’t manage what you can’t measure.” Real-time performance measurement (RTPM) solves that problem for manufacturers–that is, if they act on the data the system gives them, says John Harney
Lean initiatives with continuous improvement components have become a staple for manufacturers. This is good and bad news. Lean is a proven methodology for cutting costs—making equipment operation and business processes more efficient, making staff more capable and accountable, consolidating redundant applications—the list goes on. But lean initiatives are expensive and, if not followed up on and justified with tangible results, they can produce fool’s gold. The same is true of continuous improvement; without a performance baseline against which to gauge empirical evidence of positive change, continuous improvement becomes a guessing game based on gut instinct. There’s an old saw in carpentry—“measure twice, cut once”—that can be jiggered to fit manufacturing. With continuous improvement, that maxim transmutes to “measure continuously, adjust forever.” This is where real-time performance measurement (RTPM) is really paying off.
According to Greg Gorbach, vice president, collaborative manufacturing, at ARC Advisory Group, these services use technology that accomplishes three goals—they gather real-time and historical performance data about values like temperature, pressure and cycling from equipment and process efficiency of applications in the manufacturing plant, load it into a database to give it context related to the business goals and parameters of the manufacturing company doing the data collection, and present the data by various means (browsers, PDAs, portals) to plant and management personnel so they can use it in business intelligence and decision support applications. In short, they are the key enabling technology for continuous improvement because they measure plant operations, compare them to past performance and future goals, and present the results to the people who can make adjustments to those operations.
Most systems provide historical or real-time data or both. According to Alison Smith, senior research analyst, process and manufacturing operations, AMR Research, systems generate historical data at greater intervals—daily, weekly, etc.—while real-time data gets generated per second. That is not to say that historical data can’t be refreshed more quickly—it just typically isn’t, says Smith. Real-time data can be compared to snapshots of historical data for previous periods to compare how well equipment is running compared to satisfactory performance in the past.
These capabilities are a significant improvement over what most plants used to try to calculate and track performance in the past—Excel spreadsheets and PowerPoint for logging data that was monitored manually and not in real-time with no central repository so floor, plant management and executive personnel could not pull up data to make timely decisions about adjustments to plant operations.
But, aside from improving overall plant efficiency, RTPM is critical for other reasons. For instance, it helps manufacturers more easily comply with regulations in different industries. For instance, Hank Rossi, director of manufacturing practice at Accenture, says pharmaceutical companies “have to submit to the FDA the actual product formula and variances around each manufacturing characteristic like pressure, temperature, acidity, and the product at every step of the process has to meet the tolerances that have been set for all process characteristics.” This type of data can be measured manually, but Rossi says it’s a monumental undertaking. RTPM simplifies it.
The pioneers in this space have been around since the late-1980s and, according to Gorbach, most vendors boast four-to-six-month payback periods on installed systems, system cost is reasonable and, if the systems are implemented within a continuous improvement initiative and data is followed up on and adjustments made, then the results are dramatic, indeed. Despite these facts, Gorbach counts fewer than 10 players in this space, and estimates the total industry to be worth between $40 and $80 million. Though the industry remains in its infancy, all the signs are good that it will boom. Gorbach says it’s been one of the fastest growing markets within manufacturing, averaging about 40 percent CAGR.
ActivPlant was one of the earliest players in the RTPM space and has for some time been refining the way its technology supports continuous improvement programs like lean manufacturing, six sigma and total productive maintenance. According to Dennis Cocco, president/CEO, “you can’t drive continuous improvement unless you have a baseline of data to measure against and take a long-term view of improvement.”
Cocco says ActivPlant uses existing sensors on plant equipment to do completely automated data capture by tapping into programmable logic control signals via standard interfaces and interpreting that data into what he calls “categories of losses”—losses associated with downtime, uptime, quality and speed. Downtime, of course, measures equipment failure that interrupts manufacturing operations. Uptime measures process data associated with people. In this category, says Cocco, “equipment is capable but not currently in production because it’s waiting for parts or blocked because of downstream automation or is in the middle of a setup or changeover or not running due to lunches or breaks.” Speed measures losses in equipment cycle time when equipment is not capable of cycling at its designed rate. “If you’re losing a few seconds per cycle, that adds up over a day or week,” says Cocco. Quality measures scrapped or rejected parts.
ActivPlant also measures the activities in manufacturing processes and lets personnel retrospectively analyze each stage of a process for a particular batch of product to see where product flaws may have originated. “We look for variations in the process and how that impacts product quality and then provide complete traceability reporting,” explains Cocco. For instance, a plant might use a sophisticated computer numeric control (CNC) machining cell to cut raw metal into shape to exact specs about which you can find out more on https://parts-badger.com/rapid-prototype-cnc-machining-service-fast-parts/. If a batch of parts comes out of that process not measuring to spec, ActivPlant can take personnel back through the process and tell them which CNC cell that batch went through and what were the operating parameters such as setup dimensions, condition of the tool and machine speed to pinpoint the reason for the defect.
The systems integrate with ERP, business intelligence and maintenance management applications by means of standard interfaces and can scale across multiple plants so as to provide a common view of performance throughout a manufacturing company. The resulting benefits, says Cocco, are “better accountability for the output of the plant from plant management to senior production management to line level production management to line employees [and] bottom-line savings from reducing overtime and scrap and improving quality.” Cocco claims the products cost roughly $1,500 to $2,000 per piece of equipment, with unlimited users and data, and price-per-piece drops the more equipment that’s monitored. ROI is most definitely tied to how aggressively the manufacturer adapts its operations in response to data the systems gather.
AspenTech, another RTPM vendor, deployed one of the more impressive RTPM systems at Southern Company, the largest wholesale energy provider in the southeastern United States. Southern’s legacy environment was a mishmash of homegrown and heterogeneous vendors’ solutions that made it difficult to collect historical operational data from equipment in its 34 plants. With the AspenTech solution, via a control system, Southern now collects operational data from 250,000 sources in all plants and feeds it to a historical server. The system also transmits the information to engineers and management so they can make recommendations to floor personnel to modulate plant operations as the data indicates.
Now Southern can perform root cause analysis of equipment failure by comparing historical data feedback from periods previous to failure to periods just before. With the old systems, historical data was limited and the time stamps were not synchronized, so personnel could not get a historical snapshot of a piece of equipment. AspenTech presents a comprehensive overview of data for a piece of equipment presented in the same format over specific time periods. Southern can also do predictive maintenance for planned outages and better prevent unplanned outages. More specifically, AspenTech monitors the activity of key equipment—leaks from valves, compressor efficiency, cooling tower fan operation, and so on. The result: Southern can make better decisions about how to improve plant performance and is a more profitable operation, which helps the company provide electricity at retail prices 15 percent below the national average.