Predictive maintenance: Using AI to prevent equipment failures

Posted: September 26, 2024

My name is Petra Nieuwenhuizen, AVEVA’s Senior Product Marketing Manager for Asset Performance Management and AI. Together, I want to explore how AI-powered predictive maintenance strategies can transform how industries manage their KPIs regarding HSE, profitability, and sustainability by identifying patterns and anomalies before they lead to equipment failures.

As Lora mentioned in part two of this series, industrial AI for asset reliability and performance is a popular topic. AI helps organizations minimize downtime and therefore profoundly influences an organization’s profitability. AI for asset reliability and performance isn’t new. AVEVA’s solutions have been preventing costly downtime for decades, proving the power of AI-driven anomaly detection and empowering today’s industrial leaders.

Many brand-new tech companies and start-ups claim to have the silver bullet. They promise sky-high ROI but often don’t have the experience and track record to back it up. So rather than take my word for it, I thought I’d share real case studies from our real customers who achieved incredible individual results and ROIs, often in a single catch.

The ROI of predictive AI

It is not just about the technology. When it’s built on high-quality data and domain and industry expertise, predictive analytics doesn't just provide early warning alerts. It provides clear guidance on the forecasted time to failure and the most efficient mitigating action, and it determines the optimal time to perform the maintenance action (or the necessary production adjustments to make it to the next planned outage). In doing so, it delivers substantial financial returns, not to mention sustainability improvements. I was interested to see Lora mention skyscrapers. I immediately thought of another of the world’s tallest buildings, the PETRONAS Towers in Kuala Lumpur.

When I hear “PETRONAS,” I don’t just think of its namesake skyscraper. I also think about the impressive heights the oil and gas multinational has reached in its predictive analytics program. Using AI-enhanced analytics to optimize its asset reliability and performance, PETRONAS saved a whopping $33 million (USD) since implementing its program. PETRONAS increased its asset utilization by 0.1% per plant and reduced unplanned downtime due to 51 warnings—including 12 high-risk warnings—the equivalent of 20x ROI.

And, like Lora said, those savings were made possible by the well-rounded solution portfolio PETRONAS deployed. It had access to high-quality, reliable data through its AVEVA™ PI System™.

Or, take Duke Energy for example. With a no-code solution, Duke Energy deployed quickly and scaled effectively. In a single early-catch event, Duke Energy saved over $34 million.

Moreover, the benefits of early anomaly detection go beyond cost savings. By catching issues early, companies like PETRONAS and Duke Energy can schedule maintenance during planned downtimes to minimize disruption to their regular operations and operate most sustainably, in terms of resource and people allocation, materials, and energy use.

Key takeaways:

Real proof points, real results: Don’t fall for the AI marketing hype. Look for providers with deep domain experience and proven industry expertise.
Sustainable, actionable insights: Preventing failures before they occur doesn’t just save money, reduce downtime, and enhance operational efficiency—it’s also integral to meeting ESG goals.
Deployable, scalable solutions: Deploy fast with no-code solutions that scale seamlessly.

Predictive analytics is a game-changer for industrial operations, offering significant cost savings and operational efficiencies. “Cutting-edge AI technology” providers are a dime a dozen, nowadays, but it is not just about the technology. If you’re looking for AI-based analytics to boost your asset reliability and performance programs, you should seek out solutions that are proven, battle-tested, and industry-specific—with real case studies and results to back them up. Be sure to join us for our next roundtable discussion, in which my friend and colleague Joseph McMullen will explore the benefits of prescriptive AI.

Read part four of the series here

Series: AI for industrial asset reliability

AI’s role in industry is expanding rapidly, offering diverse benefits like improved asset performance, operational gains, and accelerated innovation. Despite its promise, skepticism persists, and many still have unanswered questions: What is industrial AI, exactly? How do I set up the right data architecture to make it work for my business? Which AI tools are best for maintaining asset reliability? And what are the first steps to implementation?

This five-part series addresses these questions, sharing insights from our experts on the evolving landscape of industrial AI and what it means for your organization.