Global Services
Interactive list componentlink componet- interactive
management, implementation of Electronic Work Instructions and optimal inclusion of additional data elements
management, implementation of Electronic Work Instructions and optimal inclusion of additional data elements
Oct 10, 2025
Author
Darren Macer
Industry
Principal Senior Technical Fellow, The Boeing Company
Welcome to the second edition of our Predictive Maintenance Newsletter
In our inaugural issue, we explored Boeing’s alerting strategy, shared insights on content development, and discussed how we measure effectiveness—along with some success stories. It stimulated engaging discussions with customers and partners at individual meetings and industry events —thank you!
Since the last installment, I had the privilege of representing Boeing at a three-day IATA-sponsored workshop that brought together operators, suppliers, and OEMs from around the world to discuss predictive maintenance. It was an enriching experience filled with presentations, networking, and collaborative discussions about our journeys and future goals. It was fantastic to see many familiar faces and meet new colleagues, including those who operate non-Boeing aircraft. This truly is a vibrant community, and we anticipate follow-up actions from the workshop, with hopes that it becomes a recurring event.
One key topic at the conference was advancing predictive maintenance from advisory insights to an integral part of the maintenance schedule. In this edition, our product team will introduce you to the Condition Based Scheduled Maintenance (CBSM) module within AHM, designed to do just that.
We’ll also take a closer look at alert utilization and the measures we’ve implemented to monitor how our alerts are being used. It’s not just about the quantity of alerts—we focus on ensuring they are accurate and delivered in a way that enables timely action. Beyond utilization, we are actively measuring alert effectiveness: when an action is taken based on an alert, we want to confirm that the alerted condition has been resolved.
On the content creation front, we’ll showcase how model-based approaches help us understand how components and systems should perform. This enables our engineers to generate insights and alerts earlier than traditional data-driven methods. This model-based health management lays the foundation for incorporating lifetime experiences—or digital twins—into predictive maintenance, a topic we’ll explore in future editions.
As part of what I expect to be a regular feature, we’ll also highlight success stories that demonstrate the real-world impact of our efforts.
On a personal note, Boeing recognizes the critical role of predictive maintenance in reducing maintenance burden. To support this, I have been promoted to Principal Senior Technical Fellow, where I will focus on maintenance vision and strategy—integrating all technical aspects to enhance operational reliability and reduce maintenance burden for both the current fleet and future platforms.
To improve operational reliability, and enhance aircraft availability, airlines are focusing on the move from unscheduled to scheduled maintenance and conducting maintenance based on known component condition. To support this transition capabilities leveraging engineering expertise and data science techniques to deliver servicing, diagnostic, and predictive alerts have been created.
~ Darren Macer, Principal Senior Technical Fellow
Benefits – Reduced labor cost and longer useful life
Condition Based Scheduled Maintenance
Traditionally, airlines have conducted manual checks at set times, which often led to unnecessary labor costs, premature component replacements, and operational disruptions, even when no issues were present. Boeing is the first in the industry to develop a regulator approved solution that shifts from manual inspections to data-driven decision-making, aimed at reducing maintenance burdens, improving operational efficiency, and enhancing cost-effectiveness in maintenance practices. The latest capability, Condition Based Scheduled Maintenance (CBSM), adheres to the MSG-3 process and incorporates Maintenance Review Board Report (MRBR) approved AHM alerts, providing a streamlined workflow for airline operators. Additionally, CBSM supports the continuous monitoring of aircraft data health, ensuring that maintenance decisions are informed and timely.
Condition Based Scheduled Maintenance (CBSM) represents a transformative opportunity for airlines, enabling a shift from traditional fixed maintenance schedules to a more efficient, condition-based strategy. By harnessing real-time data from aircraft sensors, CBSM allows operators to perform maintenance only when necessary, leading to cost savings and improved operational efficiency. The initial tasks approved for the Boeing 787, including tire pressure monitoring, brake wear assessments, and fuel filter checks, demonstrate the practical advantages of this approach. As we continue to evaluate and expand the range of tasks under CBSM, we position ourselves to further enhance maintenance practices in the aviation industry.
Journey towards an Alert Effectiveness Framework
The predictive maintenance alert lifecycle is a critical topic in the aviation industry. Predictive maintenance stakeholders, including operators, suppliers, and Original Equipment Manufacturers (OEMs), require effective frameworks to support the value proposition of predictive maintenance products and services. However, defining alert effectiveness is challenging due to
- Lack of industry standards for the end-to-end lifecycle of predictive maintenance alerts.
- Various stakeholders may want to optimize different objectives.
- Alert performance is often measured prematurely or not at all.
Data Science and older airplanes
One indication of the quality of Boeing aircraft design, is how many airplanes are flying that were first designed decades ago. The 737 had its first flight in 1967. The 747’s first flight was in 1969. While much of these aircraft have been updated and matured since then (avionics, pump design etc), there are large portions of the aircraft that are essentially the same. These systems were designed so well by engineers who have long since retired, that major portions are almost identical to the parts that first flew with that original airplane.
While this unbroken heritage is a source of pride for engineers and mechanics who work at Boeing today, it provides a unique problem for data scientists who are building modern algorithms to predict and prevent part failures that impact our customer’s reliability. The challenge for data scientists today is how to approach these heritage systems and rediscover the complex system behaviors that were designed into them years ago.
Currently Boeing is applying AI techniques to develop alert monitoring for complex systems (Example: ATA 21-Air Conditioning, ATA36-Pnuematics, and ATA 27-Flight Controls) across all Boeing aircraft. These areas are prime candidates because they're complicated, support multiple critical parts through the aircraft, and have to operate well in extreme conditions- from taxiing around a desert airport to flying near the speed of sound at heights thousands of feet higher than the peak of mount Everest). This approach of applying state of the art mathematical techniques such as machine learning, AI data mining, etc, to airplanes that have been flying since long before these techniques existed has become a core competency at Boeing.
Boeing airplanes were built to last, and modern data science approaches are built to keep them flying. I only wish that the engineers from the 1960s could see data scientists today who are bringing fresh understanding to the systems and structures they designed long before they were born.
Model-Based Health Management Approach
Moving from some of our older models to current production, and looking to the future, Boeing’s Model-Based Health Management (MBHM) approach enhances predictive maintenance by leveraging Digital System Models (DSMs) — physics-based computer models developed during the product design cycle. These models simulate how system inputs translate into outputs using physics-based relationships. The accuracy of DSMs depends on how well they represent real-world behavior. Models developed for nominal conditions (e.g., stable cruise) may not capture transient or off-nominal behaviors, which limits their predictive power. Enhancing model fidelity involves incorporating more detailed physics, nonlinearities, and operational envelope coverage.
Currently, MBHM is driving individual use cases. The 737 MAX Fan Air Modulating Valve alerts were developed using a model-based approach and the 787 predictive maintenance engineering team is developing alerts for the Ram Air Door Actuators using a model developed from the Component Maintenance Manual (CMM). Boeing’s expansion of Model-Based Systems Engineering (MBSE) practices at the enterprise level will increase DSM availability and standardization. This will enable MBHM to scale across more systems and aircraft programs, improving fleet-wide reliability and reducing maintenance costs. We look forward to updating our predictive maintenance community on advancements in MBHM as we scale the approach to more alerting cases in the future!
Thank you for your ongoing collaboration. I hope you enjoyed this edition!
Read our previous Predictive Maintenance Newsletter
Related news
