Small Steps, Big Gains in Asset Performance

By Sandy D’Souza, Director of Product Marketing at Fiix, by Rockwell Automation

World-class manufacturing is often defined by the smallest of measures. An extra few minutes of uptime on each shift could be the difference between completing an order on time or not. A slightly misaligned component can lead to a massive breakdown and hours of lost production.

Hundreds of decisions are possible for manufacturers to make every day on each asset. Each decision comes with an opportunity to make these assets, and every input that goes into its performance, more efficient. Although each opportunity may be small, it can lead to outsized value in the form of doing more, wasting less or reducing risk.

Finding and accessing these incremental gains can provide a massive competitive advantage for companies, especially in an industry where the playing field is often so level, and macroeconomic factors, such as inflation and broken supply chains, pose massive external threats.

Capitalizing on these opportunities at scale is challenging. Data is the key ingredient that makes it possible. And there lies the next great leap forward in asset performance management (APM) — using data to spot opportunities to improve asset function, then assess which opportunities bring the most value and how to execute these plans with the least amount of risk.

“The future of APM strategy is the democratization of data,” says Michael Hart, director of product strategy, Manufacturing & Industrial IoT at Rockwell Automation.

“It’s the combination of advanced algorithms and accessible, out-of-the-box systems that allow people to use these algorithms to make data-driven decisions without massive investments, learning curves and manual work.”

Big Data, Small Gains, Huge Potential

Professional auto racing offers manufacturers a glimpse of how data can be used to capture a competitive advantage through minor improvements. In the early days of the sport, teams relied on basic analytics to make decisions. Race strategies were based on data such as the ideal number of tire changes or the dimensions of the car.

The data, and how it’s used, has evolved rapidly since then. Now, teams run millions of simulations to test every possible factor before each race, from which tires to use during rainy conditions to how much fuel to use in cold weather. This is coupled with up-to-the-second data collected during a race to make real-time decisions. In a sport where milliseconds can determine a winner, these small advantages can push one team to the top.

Just as racing teams exploit every piece of data to squeeze maximum value from assets (the cars), manufacturers can do the same with their equipment. While there are often a handful of obvious ways to improve asset performance, there exists hundreds of smaller, less obvious opportunities. This ranges from cutting shift changeover times by two or three minutes to increasing raw material throughput by 1% an hour.

The changes, and their impact, are small, but the sum of each is huge.

The options also are endless. Companies that use data to uncover and prioritize these strategies will not only be the ones that succeed in good times, but also will emerge from challenging times in the best position to thrive.

Anatomy of Powerful Data

For manufacturers to run their APM program with this structure, the data they use must have three characteristics:

1. Contextualized. Data from all points in the production ecosystem are continually monitored and analyzed together so decisions can be made holistically and based on an asset’s real-time performance.
2. Integrated. Data is automatically shared and visualized between systems so trends can be identified, redundancies and inaccuracies avoided, and action can occur simultaneously at several points in the production ecosystem.
3. Applicable. Data is leveraged to make changes in real time and doesn’t require entire teams to collect, analyze, and maintain the data. This is especially important as skills gaps widen. The days of hiring entire data science teams to support an APM program are no longer feasible. They’re also not necessary. Systems are evolving to analyze data for you so it can be understood and acted on by domain experts, like plant managers and reliability engineers.

The Case for Advanced Analytics

By looking at how maintenance data can be used to find and provide small advantages with large, aggregate value, it becomes clear how this model can be applied to other areas of APM.

Maintenance Maturity Framework. The maintenance maturity framework (see diagram above) can be mapped to a manufacturer’s APM program and its attempt to go from good, to great, to world-class.

A preventive maintenance approach allows companies to make decisions about assets proactively instead of reactively. The result is less downtime, more throughput, higher output, and less waste. A predictive maintenance strategy gives manufacturers even more foresight with less manual work, multiplying the impact.

The role of data in the maintenance maturity journey and its relation to APM. Data is a connective tissue between each step in the maintenance maturity framework. For instance, to move from reactive to time-based preventive maintenance, a team might only need basic data, such as an OEM’s suggested maintenance interval.

As the team progresses to more complex strategies, they can use integrated systems to automatically pull data from several sources in real time, like meter readings and throughput. This allows the team to make maintenance decisions about assets based on what has come before, what is happening now and what will happen in the future.

Each input is a chance to make small adjustments in how an asset is managed to improve its performance. When these adjustments are applied across several assets, the performance of the entire production system is improved.

Here’s an example of what this process, and its outcomes, can look like:

• Sensors on an asset measure vibration on conveyors in real time and send this information to a computerized maintenance management system (CMMS).
• When vibration levels hit a predetermined threshold, maintenance is automatically scheduled in the CMMS for the next day. This is based on a predictive analysis of how long the asset can properly function without corrective measures.
• Bearings are replaced on the asset and the asset resumes normal operation.

The data in this example meets the characteristics set out previously:

• Contextualized: The data captures the asset’s current state and considers all relevant factors to determine the next, best action.
• Integrated: The data automatically flows from one system (equipment sensors) to another (a CMMS) in real time.
• Applicable: The data automatically triggers action in an existing maintenance workflow and any changes made from that action are documented and visualized in the system.