Digital twins can be used for a variety of applications and functions. Similar to its definition in the automotive industry, a digital twin is a virtual representation of a physical system and mimics functionalities of the actual hardware and software. In the aerospace world, digital twins provide a virtual representation of systems such as an aircraft, a satellite, or even a semiconductor subsystem within a larger system.
For the past two decades, the concept of digital twins applied only to mechanical parts and components. However, over the past four to five years, digital twins have also been used for electronic systems. This change in use is driven by the increasing complexity of compute platforms, the availability of improved capabilities and models, and the substantial amount of software run on today’s systems.
A fundamental question that teams ask when utilizing digital twins is: What system level is being considered? Is it for a subsystem like the braking system in an aircraft? Is it for a printed circuit board in the cockpit? Or is it for an underlying chip and its software that drives critical control functions with an aircraft? While the principles remain the same across different system levels, digital twins are important in aerospace and government because they can augment or replace the physical systems which were required for prototyping in the past. They are also valuable for developers in multiple locations or for demonstration to a customer or user.
For example, the Joint Strike Fighter has over 25 million lines of code and has 30,000 pounds dry weight; imagine the challenge to modify the software code and physical computer hardware as the aircraft control processing and surfaces are changed during development. The cost and time involved is enormous, let alone the possibility of achieving a healthy failure-to-success ratio. Additionally, prototyping provides the most benefit when it is utilized early in the development stage and adds significant value post hardware availability, when critical design decisions such as CI/CD flow and fault injection methodologies are being made — an impossible feat to achieve without actual hardware. This is where digital twins play a crucial role in testing, validating, and verifying both the hardware and software in real-world conditions. For example, by using a virtual manufacturing approach with Synopsys Saber™, General Motors reduced cost by 95% compared to using a physical manufacturing environment.