A Digital Twin is the digital copy of an asset’s physical representation (geometry), its behavior (simulations) and its history (telemetry). Learn more

Types of Digital Twins Peaxy offers:


A Physics-based Digital Twin (PB-DT) is a digital representation of a physical object and simulations of its physical behaviors under normal or abnormal working conditions. Learn more

The object could be a component (e.g. gear), sub-system (e.g. gearbox) or system (e.g. drive train). The PB-DT predicts physical behaviors of an object under a load. For example, it could track strains or vibrations at certain points that are measured by sensors in the field. All digital twins of a single component (such as a gear) begin with the same geometry, materials and simulations data, such as Computational Fluid Dynamics and Finite Element Analysis. A Digital Twin is unique to a specific serialized asset, and each twin is updated with specific field input data from sensors connected to a robust IoT platform.



A Multifaced Digital Twin (MF-DT) moves beyond orchestrating only physics-based models. It can orchestrate heterogeneous models and algorithms that aim to predict system-level physical and economic performance simultaneously. MF-DTs are generally applied at the system level. Learn more

While engineering teams might care more about the physical behavior of an asset, a customer might care more about the economic models an MF-DT incorporates. A wind farm builder, for example, might be concerned with the physical performance of each wind turbine, but she might also want outputs on key indicators like economic viability, cash outlays and build schedules. This information can save countless weeks, or even months of configuration work during the bidding process. The MF-DT orchestrates models and simulations across multiple disciplines on a data architecture called Aureum, which is designed to host a model orchestration environment. Aureum can manage large amounts of unstructured data, capturing the geometries for the complete facility and enabling placement on 3D maps.

Over time, field data generated by sensors and performance monitors will be fed to the MF-DT to ensure it reflects the performance and economic profile of the system while it’s in service.

What do we learn from Digital Twins?

For the customer, Digital Twins mean less unplanned down-time of equipment, cost benefits derived from condition-based maintenance regimes that rely on the PB-DT, and timely ordering of replacement parts to reduce expediting costs.

For the supplier, it means better feedback to improve and optimize designs, improved customer satisfaction and a transformed business model. Learn more

A supplier will see enhanced revenues through services and will be able to offer “equipment-as-a-service” measured against uptime metrics. Best of all, precise knowledge of a piece of equipment’s current condition makes advanced planning possible, shortens repair and inspection times, and provides necessary fleet-level updates and warranty assurances.

Physics-based Twins rely on unstructured data such as non-destructive testing data, simulations, geometry and telemetry. While indirect validation techniques can be applied, Digital Twins are validated with empirical data from testing and remote monitoring over the life of a product. Management of simulations-at-scale can be used by R&D departments to optimize designs through such approaches as Response Surface Methodology.

How will Digital Twins improve the bottom line?

Digital Twins help companies properly predict events for their assets in a way that hasn’t been possible before. Learn more

Just like their physical counterparts, Digital Twins require good “data maintenance” at the serial-number level. A “virtuous data circle” — moving from field sensors, to the Digital Twin, then back to the field through improvements in the next design cycle — is a game-changer for large industrial equipment manufacturers, suppliers and builders of large facilities.

Peaxy provides the expertise needed to build a Digital Twin as well as the hyperscale data architecture (Aureum) needed to make it work efficiently.