Monitor Quality Performance Online to Improve Customer Satisfaction

Meeting business objectives requires manufacturing managers to identify key process performance indicators, institute online measurements of those indicators, and present the results to the right people, while the product is being produced.For two decades, industrial manufacturing has looked longingly to computer-based control and automation systems as a solution to manufacturing problems.

By Dave Harrold, CONTROL ENGINEERING February 1, 2000
  • Process and Advanced Control

  • Advanced control

  • Modeling

  • Optimization

  • Productivity, Management, and Control

Meeting business objectives requires manufacturing managers to identify key process performance indicators, institute online measurements of those indicators, and present the results to the right people, while the product is being produced.

For two decades, industrial manufacturing has looked longingly to computer-based control and automation systems as a solution to manufacturing problems. Unfortunately, when manufacturing managers are asked if control and automation expenditures have met expectations, many say no.

One possible reason expectations have not been met is manufacturing managers unknowingly are part of the problem.

Manufacturing managers attend meetings where business goals, strategies, plans, and forecasts are discussed and decided. Frequently metrics established at these meetings are used to measure a manager’s performance. Unfortunately, when the manufacturing manager returns to the plant, discussions focus on scheduling and production. Unless a manufacturing manager and associated staff understand the importance of developing meaningful dynamic manufacturing performance measurements that map to strategic business objectives, disappointments may await.

Thomas Vollmann, professor at International Institute for Management Development (Lausanne, Switzerland), developed a simple model to help visualize and systematize business strategies with quantitative measurements (see Vollmann Triangle). When fully utilized, the Vollmann triangle creates a closed loop between a plant’s manufacturing strategy and the performance measurements that provide feedback on how well the strategy is being executed.

Experienced dynamic performance monitoring (DPM) practitioners advise success is best achieved when simple and direct performance measurements are directly tied to job responsibility; thus the concepts of the Vollmann triangle must be applied for each layer of responsibility (see Manufacturing Decomposition). In a multiple-layer implementation, higher-level performance measures provide targets for performance measures at the next lower level and the lower-level performance measures provide upper layers with necessary feedback.

To help users work through the complexity of developing dynamic performance metrics, major control and automation companies, as well as industry-focused consultants, offer onsite services to facilitate the process. But, as noted, many managers are wary of promised but unrealized investment benefits made in control and automation systems. Foxboro (Foxboro, Mass.) responded to skeptics with a no-capital-investment guarantee. Foxboro is so confident that implementing dynamic performance monitoring can produce at least a 200% first-year return on investment, people there have put it in writing.

Operating a manufacturing process at peak performance over long periods of time is analogous to running a high performance race car in an endurance race; it takes a lot of planning, design, attention to detail during implementation, continuous monitoring and analysis of all systems, and regular adjustments to meet ever-changing conditions.

Manufacturing managers agree meeting ever-changing customer and market demands requires high-performance race-team-like agility and flexibility, but are quick to add it is easier said than done.

Those implementing dynamic performance measurements often say it is difficult.

Complex, not difficult

Peter Martin, vp with Foxboro explains that many companies are engaged in (or were but have abandoned) various forms of Total Quality Management (TQM). Mr. Martin adds that companies who successfully implanted TQM principles, methods, and thinking into every employee’s everyday work activities have significantly benefited from their TQM investments. One of the keys to a successful TQM philosophy is ability to provide timely performance feedback. When properly implemented, DPM adds a significant element to the TQM process. Mr. Martin acknowledges implementing DPM is complex, but he claims it is not difficult and explains the difference.

“Difficult problems are scientifically challenging, but complex problems are logistically challenging. Like any complex discipline, implementing DPM requires a structured methodology that analyzes from the top down, and implements from the bottom up, similar to designing and constructing a building. An architect designs an overall building layout followed by successive layers of detail,” says Mr. Martin. Only after top-down design is complete does ground-up construction begin.

Here’s a summary of how to analyze and implement DPM:

1. Form a team that includes knowledge from the business, plant, and control domains and begin the decomposition analysis of the manufacturing facility to identify appropriate measures of performance for each operating layer. It is very important to be patient and complete decomposition to the lowest point in the structure, where an operations person has authority and responsibility.

2. Identify and rank measurements necessary to directly make, calculate, or infer the lowest level of performance measures and ensure each is intuitive to those responsible for managing the measurements. Where appropriate, absolute, rate-of-change, or deviation alarms should be assigned. It’s likely, especially in the early stages of DPM usage, that up to four performance measurements for each layer of operational responsibility will be required (see Example DPM Operator Graphic).

3. Begin implementation at the lowest operational layers and proceed upward. Implement uniformly across a layer and resolve all concerns before proceeding upward to the next layer. Depending on the manufacturing process and product complexity, there may be greater benefit in implementing all layers of one area than can be provided by all other areas combined.

4. Analyze the interaction each DPM has on one another, keeping in mind the ranking given to each. For example, experience will likely indicate a need to operate all DPMs at less than their individual optimum to maximize overall performance.

5. Initiate change-control procedures that ensure someone will manage future decisions affecting each incremental performance improvement. Having and enforcing these procedures helps ensure performance improvements remain in place.

Often the improvements gained from implementing the first few areas of the lowest layer are so great users frequently stop there. This can be a big mistake, especially long term. Users continuously attempting to improve the process are much better informed of how the process operates-what works and what doesn’t. The process knowledge gained can become a huge advantage when customer requirements or market conditions change.

Companies knowledgeable in DPM deployments are also pioneering the use of the Internet to extend e-services to users who require on-going help in optimizing manufacturing performance.

For example, Honeywell’s (Phoenix, Ariz.) Hi-Spec Solutions Uniformance suite of products and services includes Internet based-solutions that allow closer collaboration of the business, plant, and control system.

Really skeptical users might want to explore Foxboro’s recently announced risk-free onsite service offered as part of the Lifetime Performance Services suite. In addition to on-site support, Foxboro can remotely monitor DPM performance and serve as an extension to plant staff.

Today’s enlightened leaders recognize that to achieve agile and flexible manufacturing requires new management techniques. DPM is definitely part of the solution.