IOWN

Here we introduce IOWN (Innovative Optical and Wireless Network), an initiative for future communications infrastructure to create a smarter world by using cutting-edge technologies like photonics and computing technologies.

Objectives and Overview of Publishing a Reference Document to Realize Our Digital Twin Computing Initiative Together with Our Partners

The IOWN initiative aims to build a new communications infrastructure using innovative technologies that incorporate photonics in information processing infrastructure with the goal of creating a smart society through digital transformation. The three major factors for achieving this are the All-Photonics Network, the Cognitive Foundation, and Digital Twin Computing.

To first give an outline of our Digital Twin Computing (DTC) initiative, we published a white paper in 2019 describing the concept, characteristics, and potential use cases of DTC. (Currently in version 2.0) In March 2020, we launched the NTT Digital Twin Computing Research Center, which is currently in the process of holding concrete discussions together with our partners, who are experts in their fields.
Now we will provide an overview and the purpose of publishing the DTC Reference Model (English) we released on September 30, 2020, primarily focusing on Chapter 1.

Purpose of Publishing the DTC Reference Model

Conventional digital twinning, which has become widespread along with the growth of IoT (Internet of Things), is used in fields such as design, manufacturing and autonomous driving. It is used for simulations and analysis, as it perfectly mimics the shape, condition and function of things in cyberspace, reproducing them as digital twins. For example, digital twins are used in complicated airplane engine designs, for controlling robots at production sites, and more recently in controlling algorithms for autonomous driving.

Our DTC initiative takes digital twins of things and humans from a variety of industries and performs computations on them in any desired combination, with the aim of accurately reproducing combinations that could not be comprehensively handled up to now, such as humans and automobiles in a city, and to enable predictions about the future.

The key point here is that the DTC vision is not to propose specific technologies. It is an initiative geared toward creating a new digital society. In this way, it is different from conventional methods of developing technologies. We need to work with partners in a wide range of fields around the world in order to make our DTC vision a reality. To that end, the DTC Reference Model is a white paper that further defines and clarifies the concept and requirements of the collaborative framework for progressing our DTC vision.

DTC Vision: Definition of a Digital Twin under the DTC Initiative

Fig. 1 DTC vision
Fig. 1 DTC vision

Chapter 1 of the DTC Reference Model starts with describing the DTC vision. DTC includes three features: interoperability among various types of digital twins, simulation capability in cyberspace, and the DTC operations to exchange, fuse, and replicate digital twin data (Figure 1).

Although there is currently no widely recognized definition of a digital twin, the definition needs to be clarified in order to make the DTC vision a reality. Common characteristics of a digital twin proposed thus far can be summarized in two aspects: "a digital twin digitally represents a real-world entity or system," and "a digital twin enables data to be seamlessly transmitted between the physical and virtual worlds."

Fig. 2 The need for a definition of digital twin in DTC
Fig. 2 The need for a definition of digital twin in DTC

Future services that leverage digital twins will most likely utilize data from digital twins produced by different vendors. This is also the case with the DTC vision. When we came to integrate digital twins into our DTC vision, we found that previous studies were missing one important aspect: the concept of interaction between digital twins. Since many types of digital twins are combined to create a virtual society under the DTC vision, this requires various interactions among the digital twins that comprise the virtual society. Therefore, we have defined what is required of a digital twin in order to realize the DTC vision and have incorporated this into the DTC Reference Model (Figure 2).

In addition to the definition of a digital twin, this DTC Reference Model provides the structure, functions, and requirements of a digital twin in the DTC vision, in order to clarify technical aspects of a digital twin to serve as a common reference for developers, service providers, and standards organizations. As this is an architectural hypothesis of DTC, we welcome constructive comments and discussions about our proposals to promote and deepen the DTC initiative.

DTC's Four-Layered Architecture

Fig. 3 DTC's four-layered architecture
Fig. 3 DTC's four-layered architecture

The DTC white paper depicts a four-layered architecture (Figure 3). A brief outline of this architecture is provided in the DTC Reference Model.
The first layer is the Cyber/Physical Interaction Layer. This layer provides functions to collect the data required to generate digital twins through sensing real-world humans and objects, and functions to provide feedback to the real world of the results of trials and controls in a virtual society. These functions are achieved through various types of devices and algorithms such as cameras, photodetectors, light detection and ranging (LiDAR), head-mounted displays for virtual reality (VR) and augmented reality (AR), and machine learning programs.

The second layer is the Digital Twin Layer. This layer stores collected data and models that can be used to create a digital twin and provides functions to search for, update, or delete a digital twin based on its data and characteristics. These digital twins are made up of constantly updated information based on the actual corresponding objects in the real world. A digital twin stored in the Digital Twin Layer may need to adhere to a set of templates that represent each digital twin, such as an object, a computer, or a human.

The third layer is the Digital World Presentation Layer. Digital twins stored in the Digital Twin Layer will be combined and allowed to interact in the virtual society built in this layer. A virtual society can be created for specific purposes, such as a traffic environment, an urban space, and an office to simulate interactions among digital twins in a desired timeframe, location, and environment. Also in this layer, a derivative digital twin will be created that has attributes such as shape, material, and behavior that have been modified from the originals through DTC-specific operations (replication, fusion, and exchange of digital twins).

The fourth layer is the Application Layer. This layer enables applications to be implemented and executed using the Digital World Presentation Layer. We expect various types of applications to be running in this layer.

The Digital Twin Layer (second layer) and the Digital World Presentation Layer (third layer) are the main scope of this DTC Reference Model. The basic structure, necessary functions, and relevant requirements of each tier are detailed in Chapter 2 onward.

Other Key Issues

The following are the key issues featured in Chapter 2. Chapter 2 deals with the digital twin's conceptual structure. Here we assume a digital twin structure is able to handle digital twins of humans as well as objects in a unified way.

We assume a digital twin contains three types of information: intelligent, social, and physical. These three types of information further consist of data, process, and interface.
"Data" contains parameters and attributes of physical, social, and intelligent information. "Process" defines behavior of the digital twin, such as thoughts and a behavior model including autonomous decisions, especially in the case of a robot, creature, and human. "Interface" is an interface between a digital twin and other entities including other digital twins and users. We assume "Data," "Process," and "Interface" are the fundamental entities constituting a conceptual structure of a digital twin.

In addition, we have also introduced the concept of two types of digital twin structure: Primitive and Complex. Primitive is a basic element of a digital twin that does not have sub-elements. A Primitive is a single digital twin that includes physical, social, and intelligent information. On the other hand, a Complex structure has a set of Primitives and other Complexes as its sub-elements. A Complex has multiple digital twins representing features such as a machine, a building, or a human body. What is a suitable Primitive unit? There is no simple answer to this question. However, the basic guideline is that a Primitive should be of a scale suitable for observing interactions between digital twins.

Accomplishing This Together with Our Partners

Fig. 4 The main scope of the DTC Reference Model
Fig. 4 The main scope of the DTC Reference Model

As elements of creating a virtual society in accordance with the DTC vision, Chapter 3 onward explores "Digital Twin For Physical Object" (Chapter 3) and "Digital Twin for Human" (Chapter 4). Chapter 3 details the interactions between digital twins, while Chapter 4 discusses the ethical issues in changing or controlling the physical and psychological status of humans in the real world.

The current DTC Reference Model is not the final version. Following its publication, we intend to update the reference model through proposals and active discussions with our various partners in order to complete it.

References