Maintenance Mindset: It's the end of oil analysis as we know it … and I feel fine
Welcome to Maintenance Mindset, our editors’ takes on things going on in the worlds of manufacturing and asset management that deserve some extra attention. This will appear regularly in the Member’s Only section of the site. This week's column features guest contributor Michael D. Holloway, President of 5th Order Industry, writing part 2 of a series on AI and oil analysis.
The endless cycle of idea and action,
Endless invention, endless experiment,
Brings knowledge of motion, but not of stillness;
Knowledge of speech, but not of silence;
Knowledge of words, and ignorance of the Word.
All our knowledge brings us nearer to our ignorance,
All our ignorance brings us nearer to death
T.S. Eliot
A cautionary message from this quote by T.S. Eliot: a relentless pursuit of innovation, like AI for oil analysis diagnostics, must not lead to the loss of deeper meaning, wisdom, or human-centered purpose. While AI brings "knowledge of motion" through efficiency and data processing, it risks neglecting the "stillness" of reflection.
The quote warns against mistaking information for wisdom and technology for purpose. In diagnostics, this reminds us that AI should complement, not replace, the contextual judgment humans bring, ensuring technology serves humanity rather than diminishing its essence.
Industry perspective on oil condition monitoring
In order to gain industry perspective, I asked three well known pillars of oil condition monitoring "Where does AI fit into oil condition monitoring?"
Cary Forgeron, an expert in oil condition monitoring and vice president at Bureau Veritas:
"Artificial intelligence has the potential to significantly enhance the diagnosis and interpretation of data from lubricating oil analysis and condition monitoring programs. By leveraging the pattern recognition and decision-making capabilities of AI, we can improve the accuracy, efficiency, and reliability of these critical maintenance activities. At Bureau Veritas, we are actively exploring and developing AI-powered solutions to make this a reality for our clients.”
Jim Klippel, general manager of ALS Oil and Lubricants:
“Automated data processing and pattern recognition represent one of the most transformative aspects of AI in oil analysis. AI algorithms, such as machine learning (ML) and deep learning models, can process large volumes of oil analysis data, including viscosity, contamination levels, and elemental composition. These systems excel at identifying patterns or trends that might indicate wear, contamination, or lubricant degradation. ALS is looking to be the pioneer and forerunner in this work.”
Fran Christopher, U.S. director of SGS OCM North America:
“We recognize that AI tools excel at root cause analysis. By correlating oil analysis results with equipment performance metrics and operating conditions, we are looking to pinpoint the underlying causes of issues such as abrasive wear, moisture ingress, or additive functionality depletion. This capability streamlines troubleshooting efforts, allowing maintenance teams to address problems more efficiently and implement targeted solutions. SGS takes a holistic approach to the problem at hand and looks to parlay various tools to provide an increased value offering.”
Benefits and value of AI-driven oil analysis and condition monitoring
Unlike traditional analysis methods, which rely heavily on human interpretation, AI can detect subtle anomalies that may otherwise go unnoticed, enhancing both speed and reliability of the condition monitoring effort. Predictive maintenance is another key area where AI delivers substantial value. By analyzing historical and real-time oil condition data alongside operating parameters such as temperature and load, AI can predict potential failures before they occur. Early identification of warning signs like wear or contamination enables predictive maintenance, reducing unplanned downtime and extending the operational life of equipment. This approach not only saves costs but also ensures a higher degree of equipment reliability and performance.
In the past, I have been involved in automation as well as onsite sensor technology. This will work best through the integration of IoT (Internet of Things) sensors with AI, facilitating continuous real-time monitoring of oil conditions. Sensors embedded in machinery can collect data on viscosity, particle count, and other parameters, transmitting this information to AI systems for analysis. These algorithms assess the data on the fly, triggering alerts when anomalies or deviations from normal operating conditions are detected. Real-time monitoring minimizes the risk of catastrophic failures by ensuring timely intervention.
Improved accuracy and consistency in oil analysis diagnostics are another significant benefit of AI. By automating data evaluations, AI reduces human error and eliminates subjective interpretations that can vary between analysts. This consistency ensures high accuracy in identifying trends and diagnosing potential problems, resulting in more reliable maintenance decisions.
AI enhances decision support by providing actionable insights and recommendations based on oil analysis results and historical equipment performance data. AI-powered dashboards can present complex data in an easy-to-understand format, enabling technicians and engineers to make informed decisions quickly. This capability is especially valuable in high-stakes environments where timely and accurate decisions are critical.
What role will labs and technicians play?
Testing isn’t enough. SGS, ALS, and Bureau Veritas, as well as other labs, recognize that they must push further. Optimization of lubrication programs is another area where AI makes a substantial impact. By analyzing specific operating conditions and past performance data, these companies’ AI services may suggest optimal lubricant formulations or service intervals. This targeted approach improves reliability and efficiency while reducing waste and unnecessary maintenance efforts. This is where the analysis encourages bespoke formulations.
When I first began testing oil decades ago, I was taught that “a trend is my friend.” Sometimes it would take a moment or up to several minutes to recognize the trend depending upon certain variables. When you have several hundred reports to develop in a day, that adds up quickly. Trend analysis and forecasting are among AI's most powerful capabilities. By processing historical oil data, AI can forecast lubricant life, equipment wear rates, and overall system performance. This predictive insight allows for better long-term planning and resource allocation, ensuring sustained operational efficiency.
Custom solutions and training models tailored to specific industries or applications further enhance AI's utility. These systems learn from the unique operating conditions and failure modes of different equipment, making them highly adaptable and effective across various scenarios. For example, AI can be customized to address the specific needs of industries such as manufacturing, transportation, or energy.
Examples of AI applications in oil analysis are already demonstrating these benefits. In spectroscopy data analysis, AI can detect additive depletion, contamination, or wear metals more effectively than manual methods. Similarly, AI models using image recognition techniques can classify particles and contaminants such as wear debris or external particles such as dirt with exceptional precision. Machine learning algorithms also excel in anomaly detection, flagging deviations from expected patterns and enabling timely intervention. AI will bring wear debris analysis and ferrography into a new galaxy for exploration.
However, there are challenges and considerations when implementing AI in lubricating oil analysis. Data quality is a critical factor; the effectiveness of AI depends on high-quality, standardized data from sensors and laboratory analysis. Implementation costs can also be significant, though the return on investment (ROI) is typically realized through reduced turnaround time and improved reliability.
The necessity of expert oversight
It is now agreed upon that AI integrates seamlessly with predictive, preventive, and proactive maintenance programs. By analyzing oil condition data in conjunction with equipment needs, AI can prioritize and schedule maintenance tasks effectively and instantly. This integration ensures that maintenance efforts align with actual equipment requirements, minimizing disruptions and maximizing resource utilization.
So, do we need labs? Do we need technicians? Do we need diagnosticians? Expert oversight remains essential. AI should augment, not replace, human expertise. Collaboration between AI tools and experienced analysts ensures accurate diagnosis and effective action. AI is transforming the field of lubricating oil analysis and condition monitoring. By making diagnostics more proactive, efficient, and reliable, AI drives significant improvements in maintenance strategies and operational outcomes. The integration of AI not only enhances the detection and prevention of potential issues but also fosters a more sustainable and cost-effective approach to equipment management.
That's great, it starts with an oil drip
Trucks and trains, and airplanes
And JOAP is not afraid
Eye on the PM, smell how the engine burns
Maintenance serves its own needs
Don't mis-serve your KPI needs
Speed it up ten percent, speed, grunt, no, strength
The piston starts to clatter
With a fear of wrong oil pouring
Wire rope inspection when a gearbox matters,
And a consult for hire and a lock out site
Left the shop in a hurry
With the bosses breathing down your neck
Holloway's Riff of a REM Classic