Digital twins are a virtual and identical representation or model, with full characterization, of a physical product or system. A digital twin seeks to emulate the actual system by utilizing data analysis and integration, machine learning, and modeling techniques. Commercially, digital twins have been used in various applications, ranging from facility maintenance to engine modeling in order to assess system reliability and response to certain events.
Digital twins have been used across multiple industries including, but not limited to: manufacturing, product development, design customization, performance improvement, predictive maintenance, aerospace, automotive, health care, supply chain, construction, and retail. Defense and maritime applications of digital twins have been limited to optimizing shipyard operations, adding innovation, such as modernizing and standardizing IT architectures, to existing fleets, tracking ship parts, and enabling predictive maintenance of deployed assets.
Digital twins have become a critical tool and resource in allowing industries to assess, test, and analyze product and system designs prior to fabrication, instilling high confidence through data supported and integrated simulation—however, these is no formal and consistent definition of a digital twin, particularly in an undersea operational setting.
DARPA issued an Aug. 2021 solicitation for the Defining and Leveraging Digital Twins in Autonomous (DELTA) undersea operations program to help it learn how digital twins can add value to UUVs. The objective of this effort is to assess and analyze the feasibility of extending digital twins to the undersea operational environment for autonomous vehicle applications, determine the effects of undersea intermittent communications on the implementation of digital twins in this domain, and define “digital twin” use cases in an undersea operational setting.
The Navy continues to heavily invest in unmanned vehicles – including unmanned surface vessels (USV) and unmanned undersea vehicles (UUV) of all size classes – as part of a strategic shift to a more distributed fleet architecture. As such, the Navy relies on an increasingly unmanned fleet to satisfy an increasingly broad and complex set of operational requirements, particularly in the undersea domain.
This effort seeks to address whether the implementation of digital twins can add value to the UUV mission space, ranging in scale from single UUV deployments to hundreds of sorties, and if so, how digital twins should be employed in a particular scenario. In addition, this effort will investigate the feasibility of translating digital twins to the unmanned undersea environment, which has not been deeply explored by current Navy applications.
The traditional industry application of digital twins relies on robust and continuous communications with the deployed system, which are not always available in undersea operations. Therefore, determining how to address and overcome the intermittent and/or low-data rate communications challenge will be investigated.
Lastly, this effort will evaluate return on investment factors and logistical considerations associated with undersea implementation of digital twins (e.g. what size/scale of deployed UUVs is necessary to begin realizing significant operational efficiencies or advantages across the fleet, and whether digital twins should be maintained centrally at an operations center, on an individual UUV, or a combination of both). Ultimately, the vision is that digital twins could assist the operational commander in better understanding what their unmanned assets are doing, particularly in times of reduced or no communications.
There are numerous military applications for this technology. The Services are becoming more reliant on autonomous systems, and though there has been much discussion within the Department of Defense (DoD) about swarming initiatives, a real-world employment of such systems has yet to be executed and operational.
Commercial companies are already leveraging and capitalizing on digital twins and autonomous technologies. Digital twins would add a higher level of fidelity to existing models and enable real-time prediction of a system’s status, which could decrease the military’s reluctance to cede full control to autonomous agents. Current Naval and other DoD applications of Digital Twins include mirroring the activity of weapons and machines, the maintenance and upgrades of aircraft and engines, as well as government facilities.