Predictive maintenance (PdM) is a term—and a practice—you should get to know. With the Food Safety Modernization Act, or FSMA, driving new rules for food manufacturers that emphasize preventive controls, the ability to predict critical manufacturing events in your facility is going to become very important, very soon.
As the FDA shifts its focus from reaction/monitoring to prevention/enforcement, food processors will face a much higher standard for prevention. Production equipment is central to that focus. Reliability, especially the reliability of equipment that is improved through PdM, will be a big part of that challenge.
What is PdM? It’s a condition-based approach to machine reliability that identifies, measures, and earmarks factory equipment requiring maintenance before a failure occurs. PdM schedules repairs according to analysis of a machine’s health, instead of a time-based schedule.
PdM identifies issues that can’t be detected by output or visual inspection. It uses advanced diagnostic and sensing technologies including ultrasound detectors, thermography, vibration analysis, and oil analysis to identify failures at a premature phase. PdM allows maintenance to take place when it is least disruptive and most cost effective; most of all, the practice ensures that production equipment doesn’t fall below prescribed performance, quality, and safety standards.
Every major food manufacturer follows FDA Good Manufacturing Practices (GMPs) for equipment, processes, and controls. Temperature, cleanliness, and purity are all important; these factors are measured and documented on a continual basis. PdM, as a preventive discipline, supports GMP rules by identifying problems before they occur. It also helps production and maintenance personnel to address issues proactively.
Yet PdM does more than raise food safety—it makes production more efficient as well. With lean manufacturing techniques now gaining a strong foothold in food manufacturing, PdM supports such initiatives by enabling overall equipment effectiveness and total effective equipment performance. PdM, in fact, can result in both production increases and long-term savings. Research has shown that as much as 30 percent of all time-based preventive maintenance (PM) tasks can be eliminated through the use of PdM.
Putting PdM Into Effect
Implementing a PdM program, while not an insignificant task, is in many ways an extension of a more conventional PM effort. If you’re already doing some level of PM, you’ll find that a combination of PM and PdM often achieves the best results.
The first step in adopting PdM is to review and analyze your current maintenance performance. Data is a huge help here. Look at maintenance metrics like downtime, mean time between failures, parts spend, technician cost, and response time. Much of this information will be available from your computerized maintenance management system (CMMS) if your facility uses one. With this data you can determine how much economic benefit even small improvements will be worth to your company.
Next, identify the critical machines—those that, if offline, will slow or even shut down operations. Any asset that will negatively impact production, safety, and/or the environment upon failure can be considered a critical machine. Also give weight to machines that experience frequent failures, as these will likely fall into the “critical” category.
The last step in the planning process is to determine your need for critical spare parts. The goal in critical spares analysis is to maximize efficiency. If a part is readily available from vendors and not essential to equipment function, it obviously does not need to be held in inventory; one that takes weeks or months to obtain, however, may be critical in nature and thus should be kept onsite, in reserve.
Selecting a PdM Technology
With the above information in hand, it’s possible to select one or more PdM technologies for your facility. Monitoring sensors and diagnostic equipment, as well as trained technicians, represent a significant investment—so you’ll want to go with your most important technologies first.
As mentioned earlier, the most common PdM monitoring technologies are vibration analysis, infrared thermography, ultrasonic inspection, and oil analysis. In most facilities, one or two of these technologies predominate; for example, machines with fans and high-speed moving parts will benefit from vibration analysis, while electrical equipment requires advanced temperature monitoring. Failure mode and effects analysis is helpful in determining statistical norms and thresholds for machines, subsystems, and components.
Data from PdM activity must be tightly integrated into maintenance work processes. CMMS or other work order system is the appropriate destination to ensure that repair/replace work is done in a timely manner. The most advanced PdM program will be of no value if a machine experiencing imminent failure is not serviced promptly.
Finally, it’s important to track and document results. Comparing program progress against baseline data will demonstrate performance, justify cost, and uncover opportunities for further improvement. (Be sure to share results with senior management as well as maintenance and operations staffs.)
Moving from a reactive maintenance mindset to one that is predictive in nature takes time. It also requires C-level commitment since a culture shift in your processing facility will be needed. You can shorten your learning curve by researching models and practices appropriate to your situation; chances are good that no matter how unique your operation seems, someone has solved similar challenges before. Another alternative is to work with a strategic service partner skilled in PdM—someone with the necessary personnel, best practices, and technology to implement your program in the shortest possible time.
While many food manufacturers are still wrestling with the idea of predictive controls, PdM can take you a long way toward achieving the FDA’s food safety requirements. What’s more, the productivity benefits and cost savings eventually more than make up for the investment in PdM.
As you make these critical decisions, ask yourself what is the cost of inaction. Recalls are expensive—and no amount of production efficiency is worth it if the product can’t be safely consumed. Whether designed internally or strategically sourced, PdM is essential to FDA mandates in particular—and food safety in general.
Martin is director of operations, industrial parts services for Advanced Technology Services, Inc. Reach him at email@example.com.