Skip to Content

Do you really know what a digital twin is?

Michael Denis

“…when you can measure what you are speaking about…you know something about it; but when you cannot measure it…your knowledge is of a meagre and unsatisfactory kind…” Antoine-Augustin Cournot

                <p>While the statement “You cannot improve that which you cannot measure or define” is often attributed to Lord Kelvin (William Thompson) or Peter Drucker, it was Cournot, a French mathematician, physicist and economist, who first expressed this thought.</p>

The term “digital twin” suffers from a similar lack of definition and possible overuse.

People most commonly think of digital asset twins that are representations of an asset’s logical (as-designed) and physical (as-operated/as-maintained) constructs, for both structural and functional configuration of an operating asset.

There are, however, many other types of digital twins, such as digital representations of factories and assembly lines, digital representations of operating networks (e.g., air, rail, marine, transportation networks), digital representations of computing networks, digital representations of multi-echelon supply networks, digital representations of operational design risks and, most recently, digital representations of financial profitability and risk.

Servitization brings a new aspect to digital twins. Whether the term is power-by-the-hour, performance-based logistics, or tire-by-the-mile, servitization is focused on optimizing vendor profits and reducing financial risk by bundling products, services, and financing to increase asset reliability, availability, maintainability, and supportability (RAMS) at reduced operating cost. This is what some are calling a digital risk twin or digital profit twin.

In the military, this would be “increased fully mission-capable at reduced flying hours” cost or, for non-aircraft, “fully burdened total operating cost.” To accomplish this, both operational RAMS levers and inherent RAMS levers need to be addressed.

While software vendors and consultancies have been pushing model-based systems engineering (MBEngineering, MBManufacturing, and MBSustainment) and a variety of advanced artificial intelligence and machine-learning analytics, it can be easy to lose the focus on the origins of inherent RAMS. Maintenance engineering analysis and logistics support analysis capabilities are still important.

For far too long, OEMs have treated Integrated Logistics Support (ILS) as activities that have to be performed to develop maintenance programs and technical publications in order to be allowed to sell assets versus sources of improving inherent RAMS and improving servitization business models.

Ultimately, the goal is to have mature digital-analytic capabilities. The digital thread and remote condition monitoring are critical capabilities for capturing structural and functional data and delivering it to central processing nodes. AI and ML can only then provide descriptive and predictive insights into maintenance to forecast degradation or failure of components. And this should support service parts forecasting for advanced supply-chain capabilities.

The next steps are AI and ML diagnostics and task prescription to learn via digital twin causality of failure modes and degradation to determine the prescription options for various operational outcomes. Health management moves to predicting likely outcomes, given a diagnosis and prescription. The health of an asset is the delta of its physical condition to its logical or as-designed conditions. Aircraft-level prognostic health is a function of all installed systems and components.

The final step is autonomic asset performance optimization and autonomic logistics. This is self-learning, autonomous, and automatic decision support and execution capabilities from the point of operations through the entire service-support ecosystem. It simultaneously optimizes operation of an asset as well as its revenue, profit, cost, or economic performance.

For the aerospace industry, the adoption of digital asset twins, risk twins, and profit twins will depend on the maturity of an organization, but it is accelerating as companies work to address gaps while working to drive efficiencies. Digital twins are the path to long-term profitability.