Uncertainty Quantification (UQ) benefits for a Digital Engineering ecosystem

Bruce Jenkins | Ora Research

The science of Uncertainty Quantification, and the software tool SmartUQ for performing UQ and engineering analytics, were introduced in our SmartUQ: Uncertainty Quantification for more realistic engineering and systems analysis earlier this year. Today’s follow-on piece offers definitions of the concepts Digital Engineering, Digital Thread, Digital Twin and Digital Truth Source, and how Uncertainty Quantification (UQ) is crucial to successfully implementing what SmartUQ terms a “Digital Engineering ecosystem.”

These definitions come from Dr. Ed Kraft, one of the initiators and principal architects of the U.S. Air Force’s Digital Thread/Digital Twin initiative. One of Dr. Kraft’s key insights is that to successfully deploy a Digital Thread approach in support of decision making, UQ must continuously play an integral role in each stage of product development and delivery to ensure that decisions are based on reliable information.

Dr. Ed Kraft, SmartUQ Technical Advisor

Following 40 years of service, Dr. Kraft recently retired from the Air Force as Senior Leader and Technical Advisor for ground testing in the Air Force Test Center. He is now a technical adviser to SmartUQ, advising on how digital engineering and UQ can be implemented in both the private and public sectors.

Definitions

Digital Engineering—The overarching term that combines Digital Thread, Digital Twin, Internet of Things (IoT), cloud computing and big data analytics into a single ecosystem. The goal of Digital Engineering is to connect all phases of a system’s or product’s lifecycle so that each “stovepipe,” traditionally a set of disconnected or only loosely connected systems, continuously shares new insights and receives updates from a digital truth source, consisting of the collected data and models. Ultimately, digital engineering is a push from a linear, document-centric process to a dynamic, digitally-connected, and model-centric ecosystem.

“Digital engineering concepts are being adopted throughout the private and public sector, but engineering needs to transform its culture, tools, and processes to align better with a digital world,” said Dr. Kraft. “This shift to a digital ecosystem will see improvements in how products are designed, developed, produced and supported.”

Digital Thread—The digital engineering communication framework that connects all the information and knowledge across the lifecycle of a system. This challenges current practices where each “stovepipe” in the lifecycle tends to have its own models and data that are not necessarily configured or connected to others’ functional domains or the overarching system requirements.

The Digital Thread is the manifestation of the storage, retrieval, analysis and continuous redistribution of all technical data to support decisions and product development and sustainment.

Digital Twin—The parallel digital simulator to an assembled system by part number / product number, maintained and updated from design and build to operation and maintenance. If an adjustment is made to the physical system, the same adjustment is made in the digital twin for all stages of the system’s lifecycle. Big-data analytics is often applied to simultaneously understand new system data and leverage historical data for diagnostics and condition-based maintenance, and to evaluate system readiness.

Digital Truth Source—A single, digital source of information is referred to as the authoritative “truth source.” In a fully connected digital engineering environment, every user of the system and every stakeholder should have access to a single, current and configured representation of the system at a point in time aligned with current requirements to support further engineering analyses and decision making. Additionally, a digital authoritative truth source should include an assessment of uncertainty so that any additional engineering analysis or decision can take into account the reality of the system digital model and thus the probability of engineering success.

Uncertainty Quantification—The science of quantifying, characterizing, tracing and managing uncertainty in computational and real-world systems. UQ seeks to address the problems associated with incorporating real world variability and probabilistic behavior into engineering and systems analysis. More information here: https://www.smartuq.com/resources/uncertainty-quantification/

For digital engineering, UQ performs a probabilistic assessment of quantified margins and uncertainties, and is the connective tissue between model-based engineering and quantified, actionable risk assessments for decision makers. UQ ensures the digital engineering outcome will provide more affordable and better-quality products.

Available benefits

Potential benefits to product developers and stakeholders from using Digital Engineering approaches include:

  • Enhanced communication among developers and stakeholders using a single, configured, quantified source of truth.
    • Stakeholders have a window into development to more thoroughly understand the product they will receive.
    • Developers can more directly communicate their product’s features.
  • Reduced risk of not meeting performance specifications and improved product quality through continuous evaluation of requirements and design verification: Developers use a unified, consistent digital truth source with quantified margins of uncertainty which ensures robustness and avoids ambiguity in model predictions across disciplines. Requirements are digitally connected to the truth source, keeping product performance goals up to date.
  • Reduced costs and delays by using streamlined manufacturing processes with better supply chain management and less scrap and rework. The digital communication framework enables efficient coupling between design teams, customer requirements, and manufacturing production and supply chain management personnel.
  • Increased productivity due to the ability to quickly evaluate the impact of changing requirements.
  • Streamlined operations and sustainment through use of all available knowledge to optimize maintainability and extend service life of the system. Digital Engineering makes storage, access and insights gathered throughout the entire history of a product’s operation and maintenance schedule available to help engineers and product caretakers make informed decisions.

UQ benefits for Digital Engineering

UQ is a crucial component in the process of verifying reliability of the digital truth source. Decisions can only reliably be made based on information from a quantifiably verified digital truth source. UQ enables a multitude of insights that deliver a unique set of benefits within the digital engineering ecosystem:

  • Develops a disciplined, consistent approach to quantifying how well the system design meets the requirements subject to real-world variabilities by forecasting risks in domains such as product performance, cost and schedule to support better decision making.
  • Reduces maintenance downtimes and cost by enabling probabilistic analysis of the reliability of components. Maintenance is scheduled on quantified risk-based metrics instead of static requirements such as a fixed number of operational cycles.

“It is only recently that software products like SmartUQ make it practical to perform UQ analyses at each phase of the process,” said Dr. Kraft. “Integrating UQ into a Digital Engineering ecosystem connects UQ knowledge across all engineering disciplines, improving product performance as well as streamlining processes.”

UQ adoption drivers

“UQ is foundational to all engineering and decision-making activities whether digitally driven or not,” said Dr. Kraft. “UQ is a key enabler for digital engineering, but it also provides similar substantial benefit to engineering analysis in general.”

UQ has applications in many of the stages of product development; initial design and modeling, physical prototyping, testing, manufacturing, and operation and sustainment. When an engineer models a system or a component, it is important to understand the uncertainties in the outcome of the model to ensure that the model is accurate, and that a robust optimum has been achieved. Likewise, when physical experiments and tests are performed, quantified margins and uncertainties for the system are important in determining whether the system will meet requirements. In manufacturing and assembly, UQ supports statistical analysis of production processes to help control output and reduce scrap and rework. UQ also provides diagnosis and condition-based maintenance strategies for fielded systems.

Although the need for UQ in all the engineering processes has been understood for some time, most statistical UQ tools have required specialized backgrounds even to use the tool, and most engineers lack the technical depth or tools to apply statistical engineering to everyday processes, Dr. Kraft observed.

Implementation of Digital Engineering

Digital engineering has already become essential for engineering and manufacturing organizations to remain competitive. But before attempting to implement it, SmartUQ advises, “recognize that digital engineering is not an instant fix. It will require a persistent focus of corporate intent, a steady investment not only in tools and capabilities but in transforming the culture to a digital world, and a serious commitment to changing traditionally linear processes to digital approaches.”

The company sets out five steps for successfully developing a digital engineering framework:

  1. Determine the objective for implementing a digital engineering framework and what value it will bring to the organization. Use this to shape the requirements for a digital engineering ecosystem.
  2. Catalog existing capabilities that will support a digital engineering approach—analysis models, data repositories, enterprise management tools, CAD systems, manufacturing robotics and the like. Assess their utility in an integrated digital engineering environment, and any need to upgrade existing capabilities.
  3. Assess and prioritize which lifecycle processes would benefit most from a digital engineering approach. Currently, many companies have adopted digital engineering concepts for their manufacturing processes.
  4. Define the architecture for a digital engineering ecosystem to identify the software, networks and systems as well as policies and processes required to implement a digital engineering approach.
  5. Lay out the investment profile to implement the digital engineering approach to meet the requirements identified in step 1.

A number of companies have already begun implementing digital engineering concepts in their manufacturing processes. With the introduction of digital factories, manufacturers can use robotics to go from digital product models to digital manufacturing, increasing the manufacturer’s efficiency and throughput. For product support, companies have integrated digital engineering concepts with the use of sensors and IoT tools. These tools can collect and communicate the performance and health of individual products as well as collect information on their use trends to improve forthcoming products.

About Dr. Kraft

SmartUQ Technical Advisor Dr. Ed Kraft has more than 48 years’ experience in testing and evaluation in both the private and public sectors. He is one of the initiators and principal architects of the U.S. Air Force’s Digital Thread/Digital Twin initiative and a strong advocate for the application of Uncertainty Quantification in developing the digital authoritative truth source in support of decision making. A distinguished alumnus of both the University of Cincinnati and the University of Tennessee Space Institute, where he received his degrees in aerospace engineering, Dr. Kraft is a Fellow of the American Institute of Astronautics and Aeronautics (AIAA) and an Arnold Engineering Development Complex fellow.

SmartUQ