Panel Sessions and Workshop

Although the concept of predictive maintenance has been around for decades, by most accounts adoption within the process industries appears to be much slower than within other industries. In particular, the application of prognostics as an enabling technology for predictive maintenance is not widespread. This panel will aim to clarify the place of prognostics within the context of predictive maintenance, as well as address the questions regarding adoption of predictive maintenance. In particular, this panel will cover the following topics:

• Challenges with transitioning to PdM from legacy maintenance programs
• Best practices for maximizing PdM value in the process industries
• Technology gaps and horizons

• Matthew Daigle (Novity, Inc.)
• Jose Celaya (SLB)
• Ross Mahler (Moxa Americas Inc.)

Formal approaches for reporting model performance can be challenged when applied to PHM models for a variety of reasons.  Application specific factors such as characteristics of PHM data (e.g., class imbalance) combined with the highly technical nature of models may lead to performance reports which can be difficult to explain to the end user.  Further, the relationship between key decision-making factors such as economic considerations, deployment requirements, etc. may not be directly spelled out when reporting model performance.This panel will continue a discussion from the tutorial at the 2022 PHM Society conference: “Evaluating Machine Learning Models for PHM: We’re doing it wrong!”  Formal textbook model performance, common customer requirements and practical challenges such as data characteristics will be discussed.

• Jesse Williams (Global Technologies)
• Peng Liu (JMP)
• Katarina Vuckovic (Collins Aerospace)
• Dmitry Belov (SLB)

The DoE estimated that a well-designed PdM program can reduce maintenance costs by 30% Maintenance costs currently account for around 13% of total airline operation costs. The US Congress identified that “If performed effectively, predictive maintenance can reduce… system downtime, ensure adequate supply of needed parts, and decrease costs.” Army and Navy officials have identified predictive maintenance technologies as possibly preventing accidents on aircraft such as the AH-64 Apache and the F/A-18 Super Hornet and in trucking, preventative maintenance has reduced breakdowns by 20\% and maintenance costs by $2-12.5k per truck per year. NASA is working on a whitepaper on the challenges preventing the full adoption of PdM in aviation and is considering starting a body of work around this.

• Darren Macer (Senior Predictive Maintenance and Health Management Technical Fellow, Boeing)
• Justin Sindewald (Predictive Maintenance Team Lead, United Airlines)
• Rhonda Walthall (Fellow, PHM, Collins Aerospace)

Did you know that there are over 300 standards for artificial intelligence (AI) that are being developed or have been published? The intended domains include healthcare, energy, finance, manufacturing, and transportation, while the applications include computer vision and autonomous systems among others. This panel will provide information for audience members interested in learning about these standards that could be applicable to their goals, with perspectives and experiences by domain experts about standardizing AI. The panel will conclude with a facilitated discussion with the audience to explore the challenges, synergies, and opportunities of standards for AI and ideas for leadership by the PHM Society community.

• Peter Bajcsy (Project Leader, Information Technology Laboratory, National Institute of Standards and Technology)
• Kai Goebel (Director, Intelligent Systems Lab, SRI International)
• Neil Eklund (Principal Scientist, Novity)
• Xiaodong Jia (Assistant Professor, University of Cincinnati)

The emergence of ChatGPT has sparked interest in its potential impact on Prognostics and Health Management (PHM). This panel aims to explore the implications of ChatGPT for PHM, including the associated risks, how industrial companies are adopting it, and the benefits they’re experiencing. As ChatGPT gains popularity in industrial settings, understanding how it fits into PHM workflows becomes crucial, particularly concerning its capacity to optimize predictive maintenance paradigms and amplify operational efficiencies. While reports suggest productivity improvements across various industries, the exact extent and mechanisms of these enhancements remain unclear. This panel will bring together PHM stakeholders from academia and industry, and discuss topics related to integration of ChatGPT into PHM.

• Kai Goebel (Director, Intelligent Systems Lab, SRI International)
• Karl Reichard (Associate Research Professor, Graduate Program in Acoustics, The Pennsylvania State University)
• Olympia Brikis (Siemens)
• Mark Roboff (SkyThread)

The mobility sector has undergone an explosion in terms of electrification, which has brought up the need to develop PHM technologies related to power electronic devices, energy storage systems, and power transmission and usage. There is a need to explore the use of PHM in making software systems more secure and resilient. The world has seen the confluence of AI and ML, big data analytics, smart sensors supplemented by internet of things (IOT) technologies, and wireless communications technologies such as 5G and beyond networks. There is a need for a secure and resilience framework involving dev/ops with PHM to ensure continuous, uninterrupted, and reliable system operations. In this panel we plan to explore the concepts of predictive analytics, reliability, maintainability, availability, resilience, security, and system safety using a holistic PHM end to end architecture with machine learning and data management operations (DataOps) and the value of creating standards that provide a roadmap for future direction. This panel will focus on Electronics PHM Standards Development, and the coordination between two leading SDOs – IEEE and SAE – in the development of these standards. IEEE is developing a new PHM standard, IEEE P1856.1 – Recommended Practice for Development and Implementation of Prognostic and Health Management (PHM) Systems In Accordance with IEEE-1856 – which needs to be reconciled with the work of SAE committees such as E-32, HM-1, and G-34.

• Sony Matthew (SLB)
• Rex Sallade (Lockheed Martin – retired)
• Kenney Crooks (Northrop Grumman)

A Fireside Chat – Experience and Lessons Learned over the Multiple Eras of PHM Development and Implementation

This panel is made up of several “seasoned” experts who have been developing and implementing PHM related capabilities and technologies for a great number of years. This panel will use their experiences and stories to explore the issues, barriers, and lessons learned that have evolved across the many eras of PHM related activities including: requirements generation, capability benefits, ROI, justifications; development; validation & verification; policies; expanding applications; integration: implementation; operations, sustainment; enterprise-wide perspectives Please click here for an overview slide-deck for the panel. (PDF)

• Mark A. Mazarek (Boeing Global Services)
• Frank Zahiri (USAF AFMC 402d CMXG/MXDEO)
• David Peterson (Siemens)

Autonomous and unmanned systems are becoming more common in a range of civilian and defense applications. Examples include self-driving cars, mobile robots, military vehicles, and commercial transportation. Many unmanned systems are based on human-operated platforms augmented with drive-by-wire (B) and autonomy (A) kits. For operators of fleets of autonomous vehicles, the requirements and benefits of PHM are similar to those for their human-operated equivalents. However, from a sustainment standpoint, there may be additional requirements and benefits for PHM that differ from human-operated platforms. In addition, autonomous platforms have additional sensors, actuators, and software systems. This panel will explore how requirements for PHM and the implementation of PHM differ from human-operated and autonomous platforms.

• Ed Baumann (Trident Systems)
• Justinian Rosca (Siemens)
• Kai Goebel (Director, Intelligent Systems Lab, SRI International)

The PHM Society is dedicated to promoting the development, growth, and recognition of PHM as an engineering discipline and to supporting PHM education by developing standard teaching curricula in the field. The society has developed and regularly offered two courses to the community: PHM Fundamentals and Case Studies, and Analytics for PHM. It has proposed an EPD (“one-stop resource”) portal, PHM taxonomy, and PHM Professional Development Guidelines. This panel session with the full engagement of the audience aims to: summarize existing PHM educational resources; identify education and professional development needs; and prioritize new niche opportunities for society with diversity and inclusion. From this new knowledge and insights, the PHM Society will augment the EPD portal on its website with EPD resources, a prioritized list of new initiatives, and possibly create an enhanced discussion and debate area dedicated to related topics.

• Derek DeVries (Northrop Grumman)
• Peter Sandborn (CALCE)

This panel will discuss perspectives on how PHM and cybersecurity relate to and depend on each other. Main goal is to initiate discussion on the synchronous roles of Cybersecurity and PHM, and the value they add together with proactive and practical inputs. To discuss how coordinated action can lead to greater precision and higher efficiency in resolution, thus impacting the final service output for the customer.  Please click here for an overview slide-deck for the panel. (PDF)

• Bruno Paes Leao (Siemens)
• General Dave Bryan (Chairman of the Board, End-to-end Enterprise Solutions)
• Desmond Haynes (GH Konnections)

The operating fleet of nuclear power plants (NPPs) in the United States is the largest source of clean energy. Ensuring their long-term economical and safe operation is essential for the net-zero vision of the nation. Prognostics and health management (PHM) approaches are applied to enhance the situational awareness of NPP’s structures, systems, and components and achieve cost savings by automating several of the cost-prohibitive operation and maintenance activities. The application of PHM in the nuclear industry takes advantage of advancements in several technologies that include (but are not limited to) sensors, communications, machine learning, and artificial intelligence. Seamless integration of these technologies with existing infrastructure at NPPs presents several challenges that need to be understood and addressed, enabling low operating costs, high reliability and safety, and flexible plant operation and generation. The application of PHM and technologies is not just restricted to operating NPPs, but is also expected to benefit operation and maintenance activities for advanced reactors.

• Bruce Hallbert (Director of the Light Water Reactor Sustainability Program, Idaho National Lab)
• Jeffrey Clark (Constellation)
• Syed Alam (University of Chicago – Urbana Champaign)

The planned use of manned and long-term crewed space platforms, as well as quick-to-launch and reusable space vehicles, is increasing at a very accelerating rate. After the legacy NASA developed the Space Shuttle and LEO ISS; among many things, there are near-term NASA plans for a lunar Gateway station, a permanent lunar base, asteroid presence, and Mars bases. Vehicles and platforms to accomplish these far-reaching goals will include crewed space and surface-based stations and habitats; various types of launch, long-range transportation, and orbit-to-surface vehicles; and all kinds of support subsystems and technologies. Besides NASA, DoD, and other government-directed organizations; commercial-based entities are aggressively developing systems to achieve these same and additional space-related goals. These commercial-focused applications include space tourists to space and LEO, space-based hotels, and lunar and deep space resource mining. This panel will focus on issues and challenges associated with these applications; and how PHM capabilities can be applied to reduce risks, increase efficiencies, and ensure resilient sustainment of these vehicles, platforms, habitats, and systems.

• Sudipto Ghoshal (QSI)
• Health Dewey (Northrop Grumman)
• Wolfgang Fink (University of Arizona)