If run to failure is no longer an option, food and beverage processors have alternative ways to keep their equipment running at peak performance levels; they can plan maintenance schedules based on OEM suggestions. Predictive maintenance and computerized maintenance management systems (CMMSs) analyze machine components’ wear on a continuous basis and create a maintenance schedule that is proactive and at the right time, not reactive or unnecessary. Could we call it just-in-time (JIT) maintenance?
- Building in PdM
- Why PdM?
- CMMS makes order out of chaos
- Properly managed CMMS tools help ease the pain
Preventive maintenance (PM) is something we all know is necessary, but we rarely practice it with any degree of frequency. Whether it’s scheduling the car’s 3,000-mile oil change/checkup, changing the air filter in the home HVAC system on a monthly basis or purging the hot water system every year, we tend to run equipment until it lets us down unexpectedly—a maintenance approach (or lack thereof) known as “run to failure.”
With new sensing technologies coming out on a continual basis and being applied to the new equipment we purchase, we’re beginning to get warnings of imminent failures, such as your car’s “service engine soon” display or your computer’s operating system warning you to back up your data immediately due to bad sectors on your hard disk. These predictive maintenance (PdM) warnings would seem to make PM unnecessary, that maybe there’s time to plan for a failure and act upon it before a major downtime event occurs, rather than adhering to a schedule. Indeed, in light of all this new technology, you might ask, does PM even have a future?
“I wouldn’t say it’s a future, but an often overlooked and misaligned best practice,” says Greg Perry, eMaint senior consultant. “The most common mistake I see maintenance organizations do is [not to] write their PMs to the failure mode of the components.”
“Preventive maintenance is already a great step in the right direction away from reactive [breakdown] maintenance,” says Martin Brucherseifer, practice lead for asset analytics, plant data services at Siemens. Within a PdM system, permanent condition monitoring is the basis for analysis, diagnosis and the identification of trends, allowing a load-based, foresighted PdM approach. Repairs and regular maintenance tasks are performed when required and can be easily planned ahead, allowing for better budget planning.
“Preventive maintenance is just that. Many companies want to avoid downtime cost and lost production, even though there is a price tag attached [to PM],” says John Klinge, Eriez market manager-sanitary. “They have learned over time what components fail and when. But they need to strike a balance between the amount of preventive maintenance and predictive maintenance they apply in their operations. Machine builders should be integrating diagnostic technology to help [customers] easily identify components that are possibly failing, similar to how Eriez runs diagnostics for its Xtreme metal detector.”
“PM is an essential tool in any comprehensive maintenance program; otherwise, factories are continually in a reactive mode,” says Heather Betts, Advanced Technology Services (ATS) CPG site manager. “However, its use must be viewed in the context of cost and production optimization. The best programs combine PM with other maintenance tools, with PM deployed to maximize the availability of key components and subcomponents.” In addition, Betts suggests a reliability study can eliminate unnecessary PM tasks by replacing them with conditioned-based monitoring and PdM alternatives.
“One doesn’t have to look hard to find that sensors and PdM are replacing PM in all industry segments,” offers Clinton Hommel, Phoenix Contact product marketing specialist—energy monitoring & management. “The modern automobile industry is showing us that just-in-time [JIT] maintenance is the future.”
Today’s cars provide a recommended oil change based on many factors including engine idle times, ambient temperatures and driving styles. Drivers can now change the oil when it has reached the end of its life instead of replacing it at set intervals, saving the driver money and time and lessening the car’s environmental impact over the course of its life. “We are seeing similar inroads in the industry with predictive and sensor-based approaches to maintenance taking precedence over fixed PM intervals,” Hommel concludes.
“Technology doubles itself every six months,” adds eMaint’s Perry. “In turn, the accuracy of the equipment gathering the data also improves.” The true advancements are realized when tried-and-true technologies are combined to provide holistic solutions. For instance, combining wireless capabilities to automate the collection of monitored parameters takes the human factor out of the equation, he says.
“The future of PM lies in automation,” says Paul Lachance, president and CTO, Smartware Group Inc. By leveraging sensors and other data collection technologies tied with a CMMS, food manufacturers can rely on real-time equipment data points to trigger maintenance actions. For example, a PM work order or email notification (or both) can be automatically triggered if a machine produces X quantity of items or runs for X hours.
The future of PM is predictive analytics, according to David McKay, president of Wave 7/NulytIQ, a member of the Control System Integrators Association. “When predictive analytics becomes a standard, manufacturers will be able to operate further into the nominal wear zone for components and still avoid downtime. This is the Holy Grail for PM.”
“Our machines are robot based, so the information on motor/drive performance is all in a digital format and can be monitored remotely,” says Craig Souser, president of JLS Automation. “Some robot manufacturers like ABB offer remote capability so the robot/machine can advise you of a value rising past a threshold. Since you know an issue is looming, you can handle it proactively.”
For busy food processors thinking throughput rather than maintenance, arranging with the machine builder or system integrator to perform PM is another solution. “We are doing PM for a number of our customers,” says Souser. “Our equipment doesn’t need much maintenance, but it still needs an annual or semiannual checkup. These are part of our service contracts. Our staff has the expertise and experience that allow a processor to offload that work to us.”
However, a single approach to maintenance may not provide a full picture of what’s going on with equipment, especially motors. “A successful motor management plan contains three key elements: well-documented repair specifications, PM and PdM,” explains Richard deFay, project manager for Copper Development Association’s Sustainable Energy Program.1 Reliable repair specifications help ensure a motor is repaired successfully and to its maximum efficiency. PM occurs while equipment is shut down and helps keep it running; PdM occurs while the equipment is running normally and aims to identify potential problems before they occur.
Most machine builders are exploring ways to build in diagnostics and lessen the burden of malfunctions and repairs. While Phoenix Contact is primarily a component manufacturer, it offers its own solutions and engineering services, says Hommel. “A lot of the projects our solutions teams have completed within recent years involve the installation of monitoring equipment, including adding remote connectivity onto all types of systems and across many industries. Our component customers, including integrators and machine builders, are offering their [end-users] diagnostics and remote services through monitoring and connectivity solutions. Some [machine builders] include them by default; others offer them as optional functionalities.”
Are food processors looking for built-in monitoring and PdM functionality in their machines? “We hear a wide variety of comments on this, anywhere from ‘let’s be proactive’ to ‘we aren’t going to worry about it until something happens, and JLS will support it when the time comes,’” answers Souser. Whether a processor wants this functionality does not preclude built-in remote monitoring. “We can monitor performance of any system to see if it is running as it did during commissioning,” says Souser.
Eriez has incorporated sensing equipment into its new Xtreme metal detector to allow customers to monitor its health, according to Klinge. All the data is displayed in the diagnostics menu. If the sensors detect a potential failure, the user receives a warning on the HMI. The warning can be wired to the outputs to trigger an alarm light or horn. Eriez also offers 24/7 remote monitoring services in its expanding metal detection product line, which is attractive to plants with multiple machines running around the clock.
Does that mean processors are necessarily looking for a PdM environment? “To date, we have had a few requests for a PdM environment,” responds Klinge. “Preventive maintenance is still acceptable for permanent magnets and all Eriez’s core products. Our pull test kits indicate whether a magnet is performing to the highest standards. We do, however, predict companies will require monitoring of key electronic components.”
Food manufacturers may still schedule PMs based on manufacturer recommendations and their own awareness of planned maintenance and ad hoc repair histories, says Smartware’s Lachance. “A combination of approaches—automation and scheduling—is still an important component of effective maintenance operations. However, as technologies advance, pure automation will become the standard.”
“For larger rotating assets like motors, generators and gear boxes, we provide the option to build in the required sensors for condition monitoring including vibration, proximity, temperature, pressure and other sensors,” offers Siemens’ Brucherseifer. While this option may not always be economical, especially for smaller, non-mission-critical assets, including these larger assets in an overall PdM program is beneficial.
With simple assets, such as smaller pumps, fans, air handler units, etc., implementing a monthly or quarterly route-based vibration measurement is one solution. In this case, a technician uses a handheld device to measure the vibration on predefined locations of each machine that is part of the program. The technician does not need any experience in vibration monitoring if the data is transferred to a cloud-based service center with experts who do the analysis and provide insights and/or maintenance recommendations via a web portal and regular reports.
Knowing which strategies to apply and when can make a difference in quality. “The more often you cycle equipment on and off, the more you [risk] failure and [endanger] quality,” explains eMaint’s Perry. “Likewise, when deploying a maintenance strategy such as PM, you may be inducing failures by the intravenous nature of the task. PdM technology [especially in the non-destructive testing world] offers a level of condition-based maintenance [CBM] with testing done in operation under current operational conditions, hence condition-based. Controlling the condition is an important component of quality, in that it should paint the picture of how integral the right maintenance strategy is in relation to quality.”
Many automation systems, especially those in the food and beverage industry, already collect a great deal of data, adds Brucherseifer. However, much of it is not used to the extent that it could be. If data is gathered properly, experts can, through state-of-the-art data processing, correlate different types of it and inputs across an entire process to gain new insights, leading to significant improvements and optimizations. In one case, a company was able to compare two processes that were similar in nature but with very different energy consumptions. Through a detailed analysis (asset analytics), it was found one process required fewer temperature steps. By simply rearranging the sequence of the other process, the company’s efficiency was significantly improved while still achieving the same quality output.
Why would a processor want to implement a PdM environment with asset analytics? “Manufacturers have different motivations for asking us to provide asset analytics services,” says Brucherseifer. “When they want to replace a run-to-failure or PM program, the motivation is typically budget constraints. Due to unscheduled downtimes and emergency repairs, they have very high, unpredictable expenses they want to reduce and be able to better forecast.”
An efficient PdM program provides a detailed transparency about the health of each critical asset within a single or multiple plants, states Brucherseifer. Keeping the assets in proper operating condition within safe limits becomes easier and reduces incidents due to equipment failure.
“Though the P in PdM stands for predictive, there is certainly no way of truly predicting failures,” asserts eMaint’s Perry. “The best you can do is lessen the risks that contribute to failures. Upwards of 89 percent of all component failures are non-age related, with more than 14 percent related to random failures and over 68 percent attributed to infant mortality failures. In the food industry, where quality plays an extremely important part, shouldn’t the MRO [maintenance, repair, overhaul] item also have an equal level of quality assurance—assurance that comes with the right part, at the right time, in the right quality?”
Not all processors have the ability to implement PdM systems in house. “We are approached by manufacturers that already, at least partially, run a predictive maintenance program,” says Brucherseifer. “They want to extend this program to other assets, but often do not have the qualified resources to do it.” Customers like these frequently seek the support of a managed service where the expertise is provided through a strong partner company with the experience and resources to ensure reliability and continuity, he continues.
One Chicago-based CPG manufacturer with hundreds of thousands of square feet of manufacturing space outsources its maintenance to ATS, according to Betts. The manufacturer has 30 on-site technicians who respond to machine failures, assign and order parts, track work orders, access machine schematics/specifications, review histories and keep the plant up and running. The installed ATS CMMS is estimated to save each technician an hour every day in terms of paperwork and travel through the plant.
Why a CMMS? “The primary reason for a CMMS is to tell the story of maintenance, quality and safety,” says eMaint’s Perry. “So goes safety, so goes reliability. So goes quality, so goes reliability. Proactive maintenance is planned maintenance. Maintenance strategies should, therefore, be managed closely. The CMMS aids and assists by keeping what’s important in focus.”
But it can’t do this without actionable, intelligent data. What helps make a CMMS really smart is the intelligent data coming from the equipment. However, it is not just the data retrieved from a sensor; it’s also the communications and the use of multiple sensors located on the equipment. Wave 7/NulytIQ’s McKay lists the following technologies that help provide relevant data:
- The ability of modern sensors to have an IP address and be connected to a network
- Lower-costing sensors that allow a “cloud of sensors” model, providing greater granularity of data and conversion of the data into actionable information through analytics
- The combination of different sensor types on one work cell, allowing multiple analysis models.
A CMMS can monitor all this data and display it through, for example, an ATS equipment status monitor (ESM), which visually represents the plant floor, says Betts. The monitor displays the state of equipment, safety situations and PM via colors and symbols. But it’s not just a PM system. A standard feature of the ATS CMMS is its preventive/predictive maintenance planning. Data collected about work orders, parts usage, failures, etc. is used for root cause analysis (a key process in PdM), total productive maintenance and continuous improvement.
With the right data from equipment, a CMMS such as the Bigfoot version from Smartware can reduce the number of PMs and make plants more efficient. With an OPC interface, for example, machine monitoring data, such as meter readings, are transmitted to the Bigfoot system. PMs can be performed every three months instead of every four, and data transmitted through the OPC interface can trigger a real-time alert in the CMMS to notify a technician if a machine is heating too quickly or a sensor that is triggering a malfunction is a false alarm.
“Bigfoot CMMS users typically schedule PMs, trigger work orders and send out automatic reminders,” says Lachance. “Our users monitor asset costs and performance through work order history and generate reports that help with budgeting or meeting regulatory requirements, showing proof of incomplete repairs, negotiating annual contracts and/or managing or replacing capital assets.” In addition to PMs and work orders, the CMMS manages spare parts inventory. Processors can code and scan any number of parts so they can be pulled according to work orders and PM calendars. Parts can also be reserved for future work orders, and maintenance personnel can automatically refresh the available part quantity so they don’t have to keep a surplus of parts in stock.
Many processors also use the Bigfoot system for safety audits by creating PMs for safety tasks. They run reports to provide proof of completed tasks for safety inspectors, as well as PMs for keeping up to date on employee training and certifications related to worker safety, adds Lachance.
MRO items, also known as indirect materials, are fundamentally important to the world of RCM (reliability-centered maintenance), according to eMaint’s Perry. CMMS tools, such as bill of materials (BOM), PM parts, parts kitting and parts allocation, are all used in proactive maintenance planning and allow the true utilization of MRO items to stay in focus. Maintaining an accurate, true inventory listing is paramount to the mission MRO serves: the right part at the right time in the right quality or JIT.
Silver Spring Foods, a subsidiary of family-owned Huntsinger Farms, produces and bottles sauces and dressings for its own label and other national brands. Until 2010, Silver Spring’s 11 maintenance technicians used manual spreadsheets to track and manage the maintenance of the facility’s diverse equipment. But the processor obviously needed more than ERP to manage all the maintenance tasks required to keep the plant humming. The ERP system could schedule PM projects, but it wasn’t able to manage work order completion or manage the inventory of hundreds of parts to keep the coolers, mixers, pumps and other equipment running.
Prior to joining Silver Spring in 2007, Director of Engineering and Maintenance Matt Hafele knew firsthand the benefits of CMMS. After some research, he chose an eMaint X3 CMMS system to handle all the company’s maintenance tasks. While several CMMS providers charge extra for customization, Hafele found he could adapt the X3 system to his own specifications, including unique fields and descriptions, at no extra charge. eMaint University helped Hafele and his team get the system up and running quickly, importing all asset and weekly, monthly and yearly PMs into the X3 database. All work orders are classified into nine categories, from scheduled repairs to emergencies. Custom reports document how many hours the team works by activity, with each maintenance tech given access to his or her own dashboard, including tabs such as daily work order assignments.
The eMaint system has helped Silver Spring get its BRC Level A certification for food safety by providing on-demand custom reports that satisfied the audit requirements. In addition, the system schedules100 percent of the PMs, approximately 90 percent of which are completed by deadline. The system generates and tracks about 100 work orders per week.
However, when a maintenance database grows and grows without control, it may be necessary to simplify. For instance, when PepsiCo’s bottling division wanted to reduce the costs associated with the maintenance inventories at its facilities, it had to reduce the master database, which consisted of several duplicate items, obsolete materials and insufficient part descriptions. The bottler contacted IMA Limited to tackle the huge undertaking.
IMA’s data cleansing services increased the amount of useful data in the database, while reducing its size by 36 percent. The data cleansing services included vendor consolidation, part number consolidation, implementation of an aftermarket supplier program, creation of a virtual warehouse for bottling locations to share big ticket items and a supplier buy-back program.
For more information:
Martin Brucherseifer, Siemens, 678-910-9290, www.siemens.com
- “Part III: Repair Specifications, and Preventive and Predictive Maintenance,” Copper Development Association, Inc., http://www.copper.org/environment/sustainable-energy/electric-motors/case-studies/a6150.pdf.