Building a world-class predictive maintenance foundation

E. & J. Gallo’s World Class Maintenance initiative builds a predictive maintenance foundation to help process manufacturers drive improved performance.

By Tom Francisco March 20, 2024
Courtesy: Emerson


Learning Objectives

  • Recognize the impact of predictive maintenance on manufacturing efficiency and competitiveness.
  • Evaluate the role of data collection, analysis and automation in predictive maintenance strategies.
  • Learn how companies like E. & J. Gallo Winery employ automation for predictive maintenance success.

Predictive maintenance insights

  • Automation in maintenance strategies is key for optimized production. Effective data collection, contextualization and accessibility drive predictive maintenance success.
  • Seamlessly connected systems streamline data collection, analysis, and personnel notification, minimizing manual efforts and maximizing timely, accurate maintenance actions.
  • Successful predictive maintenance isn’t about indiscriminate sensor placement. It begins with critical asset assessments, pilot programs and strategic planning for effective, high-ROI implementation.

As globalization continues to connect companies with new customers around the world, process manufacturing organizations face ever-increasing competition. The resulting corporate pressure trickles down to the plant level, where new goals continuously update the speed, quality, and quantity manufacturing facilities are expected to safely maintain. To meet these goals, plants often must operate at peak production 100% of the time.

One area where many organizations are finding the key to achieving peak performance is in improved automation of maintenance and reliability strategies. Reactive maintenance is the antithesis of optimized production. When a plant cannot predict how well — or even if — its assets will run, the organization faces increased costs, delivery delays, and quality degradation. Maintenance, reliability and analytics personnel forced to perform repetitive, low-value tasks risk oversights due to human error and limitations, and they miss opportunities to perform the high-value tasks that lead directly to increased plant performance.

Today’s most successful maintenance and reliability teams are pursuing increased automation to drive their predictive maintenance strategy. Top performers like E. & J. Gallo Winery are building a strong example for others to follow. Gallo, as part of its World Class Maintenance initiative, is embracing a Boundless Automation vision for connected maintenance and reliability solutions and following up on that vision by creating a roadmap for success.

Then, they implement modern automation solutions and follow through by sharing their wins to demonstrate return on investment (ROI) to build popular support across the organization. It is a strategy any maintenance and reliability group can follow to make their team a jewel in the organization’s crown.

Building a boundless automation vision

At the heart of effective predictive maintenance is quality data. But just having data is not enough. Everyone often has more data than they can handle. To make an organization’s data work for predictive maintenance, it must be collected regularly, contextualized and made accessible by the people and systems that need it. To accomplish this, top performing maintenance and reliability teams are implementing their new technologies using a Boundless Automation vision — leveraging inherently integrated systems to move data seamlessly from the field to the edge and into the cloud.

When maintenance teams have mobility of data, they eliminate the need to spend long hours walking around plants collecting data on handheld devices, returning to the main office, uploading the data, filtering and analyzing it and manually entering it and monitoring work orders for repair.

A seamlessly connected system automatically collects data at the source, where it then automatically performs analytics at the edge or transfers critical data to more robust analytics software for analysis. When problems are identified, the software notifies the correct personnel in real time — wherever they may be.

Seamless connectivity in action

Gallo Winery delivers millions of cases of wine annually, so the company needed to find a way to streamline its maintenance and reliability practices. The organization has multiple massive facilities more than 30 miles apart. Because the facilities are so large, it was taking a very long time to complete walkarounds to perform manual rounds with handheld analyzers — a task that was performed weekly or monthly depending on the facility and equipment.

Gallo’s maintenance team found that it could not capture data as frequently as needed to properly track and trend asset health for a successful predictive maintenance program. In addition, many of the assets were very large, requiring personnel to access them from above using catwalks, and to get in tight areas below some assets to perform full assessments. As a result, some checks were missed, and the team had gaps in its data. Maintenance was further complicated by the need for the limited staff to travel among sites to ensure all assets were accounted for.

As part of an initiative to bring their organization into the top 25% of performers in their industry, Gallo’s maintenance and reliability team spearheaded an initiative to better automate their procedures to transition fully to predictive maintenance within 10 years.

The team is reducing its dependence on walkarounds and handheld monitors by installing edge analytics devices and wireless vibration monitors on rotating equipment throughout the facility. Edge analytics devices are installed on essential assets in the plant to provide actionable information right at the source. Balance-of-plant assets are candidates for the wireless vibration monitors.

All the plant sensors transmit their data to a centralized data lake managed by an enterprise-level asset performance platform that also provides data management, automated workflows, and decision support. The asset performance platform is connected to Gallo’s business system so they can automatically generate work orders and quickly complete any repairs. All work orders and parts are automatically billed to the right cost center, empowering the maintenance team to work quickly, knowing all the costs will hit the appropriate finance line.

The edge analytics devices and the machinery health software perform analytics on the data to identify the most common faults in equipment, which are then delivered directly to analysts as intuitive reports.

In most cases, the analysts can confirm the report and assign a technician, but in more complex cases they collect additional data using a handheld monitor to perform a deeper dive. In either case, automated analytics help the team cut through the noise and identify assets in need of attention.

Create the right maintenance roadmap

Few maintenance and reliability teams find success by sticking sensors on everything in the plant and hoping for the best. Successful programs are jumpstarted by thorough planning. The most successful projects often start with a pilot program, where teams identify key areas in need of improvement, and target those areas with modern automation solutions. As those pilot projects gather runtime, the team can better evaluate what works and what does not, and then adjust their larger roadmap.

Another key strategy to developing a good roadmap is to start with a criticality and problem asset assessment. Identifying the plant’s most problematic assets — even when they may not be the most critical — is often a good way to identify the ROI the organization is likely to gain by implementing a new predictive maintenance strategy. Those assets are ones the team knows will fail, so they can more quickly and clearly see the impact of catching problems early. And by following up with a criticality assessment of all the organization’s assets, the team can know where to best start allocating resources for a multi-year plan to drive the most benefit right away as they move out of the pilot stage.

Successful implementation in action

Because of Gallo’s size and the wide area where the plants are sited, the maintenance and reliability team knew they needed a well-defined plan before they started their World Class Maintenance initiative. Before the team could justify rolling out new technologies across their enterprise, they needed some reliability wins to help prove that their project would be effective.

The team started with criticality. The crushing area of operations was critical to Gallo’s busy season, and historically, the maintenance and reliability team had overhauled all the stemming machines in that area before the season began each year. Whether machines needed maintenance or not, they would be stripped down and overhauled. The process was expensive and time consuming. Moreover, opening equipment that was operating optimally also risked introducing problems where there originally were none. This area of production was the perfect candidate for a pilot program (Figure 1).

Figure 1: Gallo's stemming and crushing machines are critical to operation. Continuous monitoring helps the team keep an eye on their health at all times so they can begin the busy season with confidence.

Figure 1: Gallo’s stemming and crushing machines are critical to operation. Continuous monitoring helps the team keep an eye on their health at all times so they can begin the busy season with confidence. Courtesy: Emerson

In addition, the team added asset monitoring to a turret bearing that had been a problem asset for some time. The bearing was not only prone to failure, but also in an area that was hard to reach while equipment was running, and thus regularly overlooked. Putting an edge analytics device on the bearing provided the team not only with continuous data on the asset, but also with an up-to-the-moment health report so they could know when it was time to schedule an opportunity to shut down the equipment and intervene.

Build momentum with popular support

Building a successful predictive maintenance program is about more than catching poor asset health before it becomes a problem. The best programs thrive on popular support — from senior management and the technicians who use and maintain plant equipment every day. Documents wins is critical to the program’s success. Teams must be able to show senior management how newly installed solutions bring fast ROI, in money saved and freeing personnel for more valuable tasks.

Building momentum with technicians is often accomplished by showing them the time they can save in their walkarounds and inspections, as well as how much stress can be avoided when every repair is not a firefighting operation. While the most experienced technicians can quickly identify unusual sounds or temperatures to know something is wrong with an asset, by the time those signs are noticeable, failure has already begun. In such an event, the timeline to repair is shortened, and teams cannot schedule as easily, and repairs are more expensive — especially when a small number of specialists are centrally located and must travel to site to perform repairs.

Teams can be much more effective when predictive maintenance notifies them early. Even improvements in safety can demonstrate value to the technicians who must interact with equipment daily as part of their jobs, but they likely will not know the value unless the project team is reporting it.

Building momentum in action

At Gallo, the team scored a big win right away on their predictive maintenance project. The turret bearing—selected because it was a problem asset—presented a problem shortly after installation of the edge analytics device. The device was able to isolate the problem and narrow it to a lubrication issue, empowering the teams to quickly respond and extend the asset’s life (Figure 2).

Figure 2: Edge analytics devices process information right at the site of rotating equipment to help teams quickly diagnose the most common problems with their machines.

Figure 2: Edge analytics devices process information right at the site of rotating equipment to help teams quickly diagnose the most common problems with their machines. Courtesy: Emerson

When the team had the opportunity to stop the equipment and inspect the asset, they identified significant damage on the bearing and were able to replace it before it failed and impacted production.  Upon completion of the repair, the team immediately created a report and showed the value of the save to senior management, who saw the benefit of the program and have become among its most ardent supporters.

Gallo’s maintenance team continues to report its wins, including ones that are not as obvious, such as a more right-sized parts inventory and more accurate kitting, the latter helping crews better prepare for repairs and turnarounds.

Predictive maintenance is possible in any facility

Gallo’s World Class Maintenance initiative has delivered impressive results because, for the Gallo team, maintenance and reliability are not an afterthought. Developing highly connected automation systems helps teams at Gallo and other successful organizations better navigate their predictive maintenance efforts, ensuring that they have a constant finger on the pulse of plant health. These teams have better insight into what is wrong with each asset and can manage their resources and schedule more effectively to reduce or eliminate the reactive maintenance strategies that lead to expensive outages.

Plants run longer and faster, while their crews are safer, less stressed, and more effective. Getting there takes careful planning, but it is planning that is well within the capabilities of any organization. The efforts spent building a predictive maintenance environment today will pay dividends in competitive advantage in the coming decades.

Author Bio: Tom Francisco is a reliability subject matter expert at Emerson. He started his career in reliability with the U.S. Navy as a nuclear Machinist’s Mate and has 32 years of experience in the field. Tom graduated from University of Washington with a B.S. in mechanical engineering.