NIST数字孪生实验室报告：制造单元、数据与标准（PDF文字阅读版）
Digital Twin Lab: Supporting Research and Development for Standards and Implementations
NIST 系统集成部建立了数字孪生实验室（DTL），用于推动美国制造业采用数字孪生。报告介绍了该实验室的愿景与目标、实验室装置与已实施的案例研究，并列出实验室采用的相关标准及其在制造业数字孪生中的语境。报告还探讨了 DTL 可支撑的研究方向与未来工作。
研究报告选读。原文件共36页；这里提供可提取文字与页码链接，图中英文及扫描附录尚未转写，不能当作完整图文译本。
阅读全文（需登录） →登录一次，后续文章自动展开正文
PDF文字阅读版：原文件共36页。图片中的英文未转写，部分图像/扫描附录没有提取到正文；可通过各页链接核对原PDF。文字翻译完成不代表整份图文文件已翻译。
PDF page 1
NIST Advanced Manufacturing Series
NIST AMS 100-68
Digital Twin Lab: Supporting Research
and Development for Standards and
Implementations
Rishabh Venketesh
Deogratias Kibira
Guodong Shao
Matthew J. Triebe
This publication is available free of charge from:
https://doi.org/10.6028/NIST.AMS.100-68
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 2
NIST Advanced Manufacturing Series
NIST AMS 100-68
Digital Twin Lab: Supporting Research
and Development for Standards and
Implementations
Rishabh Venketesh
Deogratias Kibira
Guodong Shao
Matthew J. Triebe
Systems Integration Division
Engineering Laboratory
US Department of Commerce
National Institute of Standards and Technology
This publication is available free of charge from:
https://doi.org/10.6028/NIST.AMS.100-68
August 2025
U.S. Department of Commerce
Howard Lutnick, Secretary
National Institute of Standards and Technology
Craig Burkhardt, Acting Under Secretary of Commerce for Standards and Technology and Acting NIST Director
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 3
NIST AMS 100-68
August 2025
Certain equipment, instruments, software, or materials, commercial or non-commercial, are identified in this
paper in order to specify the experimental procedure adequately. Such identification does not imply
recommendation or endorsement of any product or service by NIST, nor does it imply that the materials or
equipment identified are necessarily the best available for the purpose.
NIST Technical Series Policies
Copyright, Use, and Licensing Statements
NIST Technical Series Publication Identifier Syntax
Publication History
Approved by the NIST Editorial Review Board on 2025-04-09
How to Cite this NIST Technical Series Publication
Venketesh, R., Kibira, D., Shao, G. and Triebe, M.J., (2025) Digital Twin Lab: Supporting Research and Development
for Standards and Implementations. (National Institute of Standards and Technology, Gaithersburg, MD), NIST
Advanced Manufacturing Series (AMS) NIST AMS 100-68. https://doi.org/10.6028/NIST.AMS.100-68
Author ORCID iDs
Rishabh Venketesh: 0009-0000-7521-0895
Deogratias Kibira: 0009-0002-7796-2518
Guodong Shao: 0000-0001-8856-3622
Matthew J. Triebe: 0000-0002-5964-1078
Contact Information
rishabh.venketesh@nist.gov
PDF page 4
NIST AMS 100-68
August 2025
Abstract
Digitalization in manufacturing is a concept that, while having existed for some time, can now be
realized because of advancements in IIoT (Industrial Internet of Things) enabling technologies
and data standardization. Digital twins are at the forefront of this trend and are being
implemented by many manufacturers worldwide. To further encourage the adoption of digital
twins by United States (U.S.) manufacturing enterprises, the Systems Integration Division (SID)
at the National Institute of Standards and Technology (NIST) has established a Digital Twin
Laboratory (DTL). In this report, we outline the vision and objectives of the DTL and provide
fundamental information on digital twins within the context of manufacturing. We provide a
detailed description of the DTL setup and case studies implemented in the lab. Additionally, the
relevant standards implemented in the DTL are listed and their contexts in digital twins for
manufacturing are briefly summarized. Lastly, we explore possible research areas that can be
supported by the DTL and identify future efforts to accomplish research objectives.
Keywords
data pipeline; digital twins; digital twin standards; manufacturing cell; MTConnect; small and
medium-sized enterprises; smart manufacturing
i
PDF page 5
NIST AMS 100-68
August 2025
Table of Contents
1. Introduction ...................................................................................................................................1
1.2.1. Testing and Developing Relevant Standards ................................................................................ 2
1.2.2. Prototyping Digital Twin Systems and Components ...................................................................... 2
1.2.3. Testing Enabling Technologies ..................................................................................................... 3
1.2.4. Providing Demonstrations and Enabling Collaborations on Digital Twins ..................................... 3
1.2.5. Developing Tools for Digital Twin Prototypes ............................................................................... 4
2. Digital Twin Lab Operating Environment and Equipment .................................................................4
2.1.1. Manufacturing Work Cell Layout and Material Flow ..................................................................... 4
2.1.2. Communication Network ............................................................................................................. 5
2.2.1. PocketNC ...................................................................................................................................... 6
2.2.2. Automated Fixturing Solution ...................................................................................................... 7
3. Digital Twin Lab Data Infrastructure ................................................................................................9
3.1.1. Data Collection from Equipment ................................................................................................ 10
3.1.2. Data Aggregation ........................................................................................................................ 10
3.2.1. Data Streaming Requirements ................................................................................................... 11
3.2.2. Technical Implementation .......................................................................................................... 11
3.3.1. Data Storage Requirements ....................................................................................................... 11
3.3.2. Technical Implementation .......................................................................................................... 12
4. Relevant Standards and Implementation ...................................................................................... 12
5. Research and Development Efforts ............................................................................................... 13
5.2.1. Collaborative Robot Digital Twin ................................................................................................ 14
5.2.2. CNC Machine Tool Digital Twin .................................................................................................. 15
ii
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 6
NIST AMS 100-68
August 2025
6. Conclusion and Future Work ......................................................................................................... 16
References ....................................................................................................................................... 17
Appendix A. MTConnect Device Configuration File ............................................................................. 19
Appendix B. MTConnect Agent Configuration File ............................................................................. 29
List of Figures
Fig. 1. Material Flow of the DTL Manufacturing Work Cell ...................................................................5
Fig. 2. Material Flow of the DTL Manufacturing Work Cell ...................................................................6
Fig. 3. Axes of the Pocket NC ..............................................................................................................7
Fig. 4. Expanded CAD model of zero-point fixturing solution for Pocket NC machine tool .....................8
Fig. 5. Data pipeline diagram ..............................................................................................................9
Fig. 6. Part design to be cut on the CNC, handled by the robots, and inspected by the CMM .............. 13
Fig. 7. In progress robot digital twin for model validation and improvement ...................................... 14
Fig. 8. Altair simulation of CNC machine tool .................................................................................... 15
iii
PDF page 7
NIST AMS 100-68
August 2025
Acknowledgments
We greatly acknowledge Yehoshua Halle, Eric Charlery, William Stiller, Aubrey Simonson, and
Laetitia Monnier for their work in developing digital twins and digital threads for the lab,
Michael Chen for his work in setting up the data infrastructure, and Simon Frechette and Josh
Lubell for their support on the hazard review of the lab.
iv
PDF page 8
NIST AMS 100-68
August 2025
1. Introduction
In the United States (U.S.), most, if not all, large manufacturing companies have made digital
manufacturing a priority and started to implement digital twins in their facilities and supply
chains. Although the relevant technologies for digital twins originated in the 1960’s with the
push for space exploration, the term was not coined until 2002 by Michael Greaves – an expert
in product lifecycle management [1]. In this report, we focus on digital twins for
manufacturing. The role of digital twins in modern manufacturing is so well recognized that
the most recent U.S. National Strategy for Advanced Manufacturing calls for efforts to
“Enable the application of advanced sensing control technologies and machine
learning across the manufacturing sector. Advance smart manufacturing by pursuing
digital twins. Develop standards for data compatibility to enable seamless integration
of smart manufacturing.” [2].
In response, the National Institute of Standards and Technology (NIST) established a Digital
Twin Laboratory (DTL) to advance the development and adoption of standards to support the
use of digital twin technology for manufacturing.
A barrier to advancing digital twin technology in manufacturing is the need for commonly
accepted definitions for the concept and relevant terms. Aspects of digital twins such as
accuracy, synchronization, and bidirectional communication with real systems and their
significance are all debated topics that make defining digital twins a not-so-simple task. This
report adopts the terminology and structure defined in the ISO 23247 (Digital Twin
Framework for Manufacturing) standard to provide a reference framework for digital twins in
manufacturing. This framework includes the following definitions of manufacturing digital
twins and relevant terms [3].
• manufacturing digital twin - “a fit-for-purpose digital representation of an
observable manufacturing element with synchronization between the element
and its digital representation”
• observable manufacturing element - “an item that has an observable physical
presence or operation in manufacturing”
• digital representation - “a data element representing a set of properties of an
observable manufacturing element”
This report discusses the vision and objectives of the DTL (Section 1), describes the hardware
and relevant setups of the DTL work cell, and provides information on the data pipeline
implemented in the DTL (Sections 2 and 3 respectively). In addition, it describes ongoing
standards work and research efforts that leverage the DTL (Sections 4 and 5).
1
PDF page 9
NIST AMS 100-68
August 2025
Lab Vision
The vision of the DTL is to enable innovation by advancing measurement science, standards,
and technology in the field of digital twins for manufacturing. The DTL utilizes critical
technologies to test and promote the use of relevant standards for U.S. manufacturers to
successfully implement digital twins in their operations.
Lab Purpose and Objectives
The purpose of the DTL is to serve as a testbed to support research areas pertaining to digital
twins in manufacturing. The DTL supports five primary objectives.
• Testing and Developing Relevant Standards
• Prototyping Digital Twin Systems and Components
• Testing Enabling Technologies
• Providing Demonstrations and Enabling Collaborations on Digital Twins
• Developing Tools for Digital Twin Prototypes
This section discusses these objectives and how the DTL supports each one.
1.2.1. Testing and Developing Relevant Standards
To promote the adoption of digital twin technologies, identifying and developing relevant
standards is necessary. Standards allow for interoperability between enabling technologies
(e.g., IIoT, smart sensors, communication protocols, modeling tools, and visualization
applications) and various manufacturing tools in today’s market. Additionally, by providing
common language, standards support industry partners in communicating and sharing their
perspectives on digital twins for manufacturing. The DTL provides a platform to implement
standards such as ISO 23247 and MTConnect [4] [5] [6] [7] [8]. Through implementation,
collaborators can demonstrate the advantages of leveraging existing standards, identify areas
of improvement in said standards, and provide feedback to the relevant standards
organizations.
One example of leveraging the DTL to test standards is the implementation of ISO 23247 in the
lab. This standard outlines the requirements of the digital twin framework for manufacturing.
Section 5.2 provides concrete examples of how the framework is implemented and tested.
Another example is the use of the MTConnect Standard in the DTL data pipeline (see Section
3). Through collaboration with the MTConnect Institute, the lab supports the testing and
development of their latest revision for the MTConnect standard. More information on
relevant standards can be found in Section 4.
1.2.2. Prototyping Digital Twin Systems and Components
At their core, digital twins are complex systems comprised of multiple components, e.g., data
2
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 10
NIST AMS 100-68
August 2025
collection, data modeling, and user interfaces. The hardware and software infrastructure of
the DTL facilitates the prototyping of individual digital twin components and integrating them
to form complete digital twins. Key digital twin components that can be developed through
prototyping in the DTL include components for (a) data collection and communication and (b)
data management. For the former, the DTL supports the implementation of multiple
industrial communication protocols such as Modbus TCP/IP in combination with semantic
standards for manufacturing data such as MTConnect. For the latter, the DTL supports the
implementation of data management solutions, such as MongoDB (a No-SQL database
product) to store and query historical datasets. Prototyping digital twin systems and
components allows for the testing of standards and exploring scenarios where digital twin
components can be reused for different use cases.
1.2.3. Testing Enabling Technologies
A driving force behind the rise of digital twins in manufacturing is the advancement of enabling
technologies. These technologies include, but are not limited to, sensors, automation
technologies, modeling and visualization tools, extended reality (XR) [5], simulation tools, and
up-to-date data management solutions. Since digital twins operate at the intersection of these
technologies, testing and validating these technologies is critical to success. Not only does the
testing instill confidence in stakeholders looking to digitalize their shop floor, but it also helps
them apply relevant standards (see Section 4), develop frameworks, and implement use case
scenarios. Due to its reconfigurability and flexibility, the DTL provides an opportunity to deploy
various enabling technologies. An example of leveraging enabling technologies to enhance a
work cell is an automated material fixturing solution being implemented on the Pocket NC as
described in Section 2.2.2 to support research in automated workflows for manufacturing cells.
1.2.4. Providing Demonstrations and Enabling Collaborations on Digital Twins
As a managed source of data and a testbed supporting different work cell configurations, the
DTL offers opportunities for both internal and external collaborations. These opportunities
include (a) working with machine-tool vendors to implement capabilities supported by digital
twin standards and development (e.g. implementing MTConnect Part 5 - Interfaces on a
Mitutoyo Coordinate Measuring Machine), (b) optimizing manufacturing processes to reduce
waste and maximize productivity (e.g. exploring the intersection of digital twins and LEAN
manufacturing principles), and (c) working with Standard Development Organizations (SDOs)
such as the MTConnect Institute to test revisions to standards. In addition to collaboration and
research, the DTL demonstrates how digital twins for manufacturing applications can be
developed and deployed. Furthermore, digital twin use cases give both internal researchers and
external stakeholders an understanding of the capabilities of digital twin technology when
paired with appropriate data visualization tools.
3
PDF page 11
NIST AMS 100-68
August 2025
1.2.5. Developing Tools for Digital Twin Prototypes
The set of software tools available in the lab range in capabilities such as collecting and
transforming raw data from equipment, aggregating data from multiple equipment and certain
stages of a product’s lifecycle (design, manufacturing, and inspection), converting between
multiple data standards, and interfacing with client-side applications (e.g. simulation modeling
tools, database solutions, augmented reality, virtual reality, etc.) [4] [5]. While some of these
tools are publicly available many of the tools were developed in-house as part of research in the
DTL. Stakeholders can use our library of tools to replicate the setup in the DTL or adapt to
similar use-cases.
2. Digital Twin Lab Operating Environment and Equipment
The DTL contains equipment and hardware, including collaborative robots, a machine tool,
inspection equipment, and computing hardware. All equipment is connected centrally to
support the data infrastructure described in Section 3. This section (a) describes the hardware
and software setups of the DTL and how the current lab configuration supports the flow of data
from machines to client-side applications (i.e. data pipeline), and (b) explains the capabilities of
each piece of hardware and respective setups.
Manufacturing Work Cell and Communication Network
Cellular manufacturing is a “lean method of producing similar products using cells” to optimize
operations and more efficiently utilize resources [9]. Some advantages of cellular manufacturing
include work-in-progress (WIP) reduction, space utilization optimization, lead time reduction,
increased productivity, and greater flexibility [10]. To remain competitive, U.S. manufacturing
enterprises often adopt this style of manufacturing. However, challenges such as balancing
cost, efficiency, and flexibility still exist. The DTL aims to support research in cellular
manufacturing through a small-scale manufacturing work cell that supports digital twinning
with full automation for various use cases.
2.1.1. Manufacturing Work Cell Layout and Material Flow
Cellular manufacturing is a “lean method of producing similar products using cells” to optimize
operations and more efficiently utilize resources [9]. Some advantages of cellular manufacturing
include work-in-progress (WIP) reduction, space utilization optimization, lead time reduction,
increased productivity, and greater flexibility [10]. To remain competitive, U.S. manufacturing
enterprises often adopt this style of manufacturing. However, challenges such as balancing
cost, efficiency, and flexibility still exist. The DTL aims to support research in cellular
manufacturing through a small-scale manufacturing work cell that supports digital twinning
with full automation for various use cases.
4
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 12
NIST AMS 100-68
August 2025
Fig. 1. Material Flow of the DTL Manufacturing Work Cell
While Fig. 1 describes the current setup of the manufacturing cell, the DTL is reconfigurable,
thus supporting the implementation of multiple material flow and layouts.
2.1.2. Communication Network
An important consideration when configuring the hardware in the lab is to support the
development of a flexible data pipeline that allows data to be collected from each stage of the
manufacturing process from production to inspection. Figure 2 shows the network diagram
that describes the manufacturing equipment layout and computing resources. The
collaborative robots, the CMM, and the CNC are directly wired to an ethernet switch. This
switch then connects directly to the lab’s wireless router. A Linux computer that hosts the
MTConnect Agent (see Section 3.1) connects to the router as well. This enables a local subnet
in which all the manufacturing equipment communicates with both a central data aggregator
and each other. Database services are also deployed on a Linux computer. Lastly, a Windows
computer wirelessly connects to the router to pull data from the MTConnect Agent. The
Windows computer has simulation modeling tools (e.g. Altair and MATLAB Simulink) that
researchers can use for digital twin research and demonstrations.
5
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 13
NIST AMS 100-68
August 2025
Fig. 2. Material Flow of the DTL Manufacturing Work Cell
Computerized Numerical Control (CNC) Machine Tool
The CNC machine tool used in the lab is a manufacturing technology that utilizes subtractive
manufacturing to produce designed parts and is the core of our work cell. Additionally, the CNC
machine tool provides numerous data items that include operational data, machine status data,
and data for setups and tool path data. This section briefly describes the CNC machine tool used
in the lab and the extra hardware deployed.
2.2.1. PocketNC
The work cell uses a Pocket NC V2-10, a 5-axis desktop milling machine. With 5-axis milling
capabilities, the lab can simulate the operations of larger industry counterparts for
subtractive manufacturing. This mill has a 50,000 RPM spindle and travels 115.5mm (X), 128.3
mm (Y), and 90.1mm (Z). The table rotates from -25° to 135° in the A direction while the table
can continuously rotate in the B direction (-9999° to 9999°). These axes can be seen in Fig. 3.
The Pocket NC operates with most CAM software. The machine accepts standard G-code, which
enables the results to be scaled up to large workshop-size machine tools. Desktop-scale
manufacturing equipment makes it practical to provide researchers with accessible equipment
6
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 14
NIST AMS 100-68
August 2025
whose results are comparable to those of workshop counterparts.
Fig. 3. Axes of the Pocket NC
2.2.2. Automated Fixturing Solution
An automated material fixturing solution for the Pocket NC supports full automation of the
work cell. While multiple automated material fixturing, solutions are available for traditional
shop floor machine tools, the small size of the Pocket NC necessitates a custom solution for
the machine tool. A zero-point fixture (a fixturing system that uses pre-defined reference
points) solution ensures there is only one possible point to fixture parts, allowing for higher
repeatability. This is especially important for work cell automation, as even the slightest
deviation in the material fixture can lead to significant discrepancies in part quality.
Figure 4 shows an expanded CAD model of the zero-point fixturing system. A custom-built
adapter connects the clamping module to the Pocket NC’s B-Table (see Section 2.2.1). The
adapter serves as a connector between the clamp and the machine tool, and it also features air
hose fittings to power the air-actuated clamping module. While not shown in Fig. 4, the
fixturing system requires the installation of a push-to-connect swivel connector on the base of
the adapter for the air hose. Schunk’s Vero-S module serves as the air-actuated clamping
module. It requires at least six bars (600 kPa) of actuating air pressure to be supplied to release
the clamping mechanisms. Lastly, to attach a workpiece to the clamping system, two steps must
be completed to prepare stock material prior to clamping. The first is installing a clamping pin
7
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 15
NIST AMS 100-68
August 2025
to each piece of stock. The clamping pin clamps onto the Vero-S module, and the pins are
tailor-made for the Vero-S module. The second step is to machine a notch in the stock material
so that the workpiece is rotationally locked with the clamping module, as shown in Fig. 4.
Fig. 4. Expanded CAD model of zero-point fixturing solution for Pocket NC machine tool
Collaborative Robots
The DTL uses UR5e robots that are lightweight (20.6 kg) and maneuver with six degrees of
freedom. They have a maximum payload of 5 kg, a reach of 0.85 m, and a repeatability of ±
0.03mm. The UR5e robots are used for material handling activities in the work cell: picking
parts from the input storage, loading onto the Pocket NC, offloading the Pocket NC, loading
the CMM, and offloading the CMM to the output storage. These robots are programmed by
either writing a program in URScript and loading it to the controller or using a teach pendant.
The robot end effector is the 2F-85 Gripper and is designed for material handling operations.
The gripper has two articulated fingers with two joints for each finger. These features enable it
to handle parts of various geometries required for pick and place, machine tending, assembly,
and quality testing. The gripper can be used in both time-based control and event-based
control schemes. Like the UR5e robots, the gripper can be programmed by either writing a
program and loading it to the controller or using a teach pendant.
Coordinate Measuring Machine
A CMM is used to measure the geometrical features of a physical object by touching the
surfaces of the object with a probe to measure discreet points. The CMM supports product
quality conformance by validating that the product matches the design. When a part is
measured by a CMM, data regarding the dimensions of the part geometries are collected and
analyzed to determine if there are any nonconformities. The results of this analysis provide
feedback to the manufacturing system or process [12]. Excessive geometric variation in the
finished product can then be avoided.
8
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 16
NIST AMS 100-68
August 2025
The CMM in the DTL is a Mitutoyo Ko-ga-me that can characterize and inspect parts manually or
automatically with a part program. This CMM can be integrated with other equipment in the
work cell, such as the CNC machine tool and robot arms. This integration provides real-time
feedback is essential for effective process monitoring and quality control, as opposed to mere
inspection in a standalone mode.
3. Digital Twin Lab Data Infrastructure
The DTL provides a wealth of data ranging from operational/status data from machines and
metrology data from product inspection. This section describes the data infrastructure of the
DTL. Additionally, this section discusses relevant standards and tools to provide background on
the data pipeline’s setup and implementation.
MTConnect Implementation
This section briefly overviews the MTConnect standard and explains how the standard provides
a foundation for the DTL data infrastructure. Figure 5 shows the flow of data in the data
pipeline. Each aspect of the data pipeline is discussed below.
Fig. 5. Data pipeline diagram
The MTConnect standard provides a semantic vocabulary for manufacturing data that defines
structure and context. The standard serves to facilitate interoperability between data coming
from various machine types and different vendors. In addition to the semantic vocabulary
offered by the MTConnect standard, the MTConnect Institute provides a set of tools that allow
the collection of raw data from machines, the transformation of the data to an MTConnect-
9
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 17
NIST AMS 100-68
August 2025
compliant data format, the aggregation of data from multiple machine tools and sensors, and
the supply of data for client-side/third-party applications. For more information on the
MTConnect standard and how the available tools operate, please refer to the NIST Advanced
Manufacturing Series Report for “Connecting, Deploying, and Using the Smart Manufacturing
Systems Test Bed” (NIST AMS 200-2) [13] and the MTConnect standard documentation
[14][17].
3.1.1. Data Collection from Equipment
All the DTL equipment (UR5e robot arms, Pocket NC, and Mitutoyo Ko-ga-me CMM) can
generate operational and condition data. In addition, the CMM provides inspection data.
Appendix A shows the device configuration file for the data items from each machine while
relating proprietary data tags to MTConnect data tags. For a list of MTConnect Data Tags
with their respective contextual information, refer to Part 2.0 of the MTConnect Standard –
the Device Information Model [14]. For each machine, software adapters transform raw
data into a MTConnect-compliant data format. The universal robot controllers do not
support the installation of third-party software. Therefore, Raspberry Pi’s collect raw data
from the UR5e’s via ethernet and have the adapters installed on them. The MTConnect-
compliant data format is known as SHDR, which is a pipe-delimited text format structured
as follows:
YYYY-MM-DDTHH:MM:SS.SSSSZ|DataTag|DataValue.
The SHDR format associates a data tag and data value with a timestamp in UTC (Coordinated
Universal Time). Also, of note, a Quality Information Framework (QIF) document, for which
MTConnect provides a wrapper in native eXtensible Markup Language (XML) format, is used to
aggregate and organize inspection data from the CMM.
3.1.2. Data Aggregation
For data aggregation purposes, the MTConnect C++ Agent (v 2.1) is hosted on the Linux
computer, referred to in section 2.1.2. Upon transforming raw machine data to SHDR, the
adapters for each machine send the SHDR data to the Linux host via socket-to-socket
programming. Upon receiving data from the adapters, the Agent validates the data to ensure
that the data coming in adheres to the MTConnect standard. The redacted agent configuration
file with local network information can be found in Appendix B.
Data Streaming
The DTL offers different data streaming capabilities via a data pipeline. The data pipeline lays
out how to manage a constant flow of data from disparate sources and supports leveraging the
DTL for different use cases. The MTConnect Agent is a software tool that collects and organizes
MTConnect data and allows us to implement the standard in our data pipeline.
10
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 18
NIST AMS 100-68
August 2025
3.2.1. Data Streaming Requirements
The requirements for data streaming are as follows:
a. the data stream should be readily available to internal researchers for integration
with client-side applications
b. the data should have context associated with it so that other stakeholders can
better understand the dataset
c. the data stream should support addition of more manufacturing elements (machine
tools, sensors, etc.)
researchers should have the ability to query based on specific components, devices, and time
intervals.
3.2.2. Technical Implementation
The MTConnect Agent is a software application provided by the MTConnect Institute that
collects machine data, organizes the data, and provides the data in a MTConnect-compliant
format. The Agent also receives and processes data requests from client-side applications.
The MTConnect Agent, upon aggregating and validating data from the machine adapters,
serves up the data in a machine-readable format – namely XML. Users can query the agent
based on device name, component name, and sample size using the Hypertext Transfer
Protocol (HTTP) requests. (As of this publication, the Agent is only available internally at
NIST.) Upon querying the data, researchers obtain an XML document with their dataset that
can be parsed and used for third-party applications. It is important to note that the Agent has
a max buffer size of 131,072 samples, thus making it unsuitable for storing historical data
over significant intervals of time. Therefore, a data storage strategy is in place as described in
section 3.3.
Data Storage
Application use cases in the DTL may require the use of historical datasets, and a data storage
strategy enables the DTL to serve as a managed source of data for various use cases. The data
storage strategy is driven by the requirements of our setup and use cases but can be modified
for similar digital twin implementations.
3.3.1. Data Storage Requirements
The requirements for data storage are as follows:
a. the managed data should be readily available to internal researchers for integration
with client-side applications
b. users should have ease of access to datasets without technical knowledge of the
data storage implementation
11
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 19
NIST AMS 100-68
August 2025
c. the data storage solution should have the sufficient capacity and ability to
archive historical data sets
researchers should have the ability to query based on specific components, devices, and time
intervals.
3.3.2. Technical Implementation
A MongoDB implementation handles the data storage in the DTL. MongoDB is an open-
source No-SQL Database solution. No-SQL (Structured Query Language) solutions are non-
relational databases. Compared to relational databases (SQL), non-relational databases
prioritize rapid development and accommodate the addition of various data attributes
throughout their usage.
The implemented methodology is as follows. The MTConnect Agent is routinely sampled (~1 Hz)
using HTTP requests that result in XML documents containing data from all the machines over
the sampling period. These XML documents are then parsed, and the data are restructured to
be added to the MongoDB instance. Additionally, for user convenience, a Graphical User
Interface (GUI) helps users generate MongoDB queries.
4. Relevant Standards and Implementation
Digital twin standards, such as architectural frameworks, can provide implementation
guidelines for the development of digital twins. Specific standards support digital twin
functionalities such as data collection, communication, information modeling, systems
integration, system abstraction/modeling, automation, and control. The set of standards
currently being implemented in various digital twin components in the DTL are described
below. New applicable standards will be continuously added to the list in future versions of
this report.
• ISO 23247, Digital Twin Framework for Manufacturing, provides a generic guideline,
a reference architecture, and a framework for case-specific digital twin
implementations. The standard supports the composability of models and
interoperability among digital twin modules. It also provides examples of data
collection, data communication, integration, modeling, and applications of relevant
standards [15]. ISO 23247 provides the framework for all the digital twin
development in the DTL (see Section 5) and the concepts and definitions from ISO
23247 are adapted to this paper.
• ISO 10303, Automation Systems and Integration – Product Data Representation
and Exchange, also known as Standard for the Exchange of Product (STEP) model
data, supports the exchange of product manufacturing information [16]. We
implement ISO 10303 during the product design process (see Section 5.1) and
when modeling lab equipment for digital twin visualization (see Section 5.2).
• The American National Standards Institute (ANSI) MTConnect supports
interoperability by providing a vocabulary for manufacturing equipment, making
12
PDF page 20
NIST AMS 100-68
August 2025
possible structured contextualized data, and avoiding proprietary formats. Data
sources of MTConnect in production include equipment, sensor packages, and other
factory-floor hardware [17]. MTConnect is the foundation for the DTL’s data
pipeline (see Section 3) and the MTConnect Agent implemented in the pipeline is a
software tool provided by the MTConnect Institute.
QIF is a framework standard for computer-aided quality QIF systems. QIF enables the capture,
use, and re-use of metrology-related information throughout the Product Lifecycle
Management (PLM) and Product Data Management (PDM) domains. It supports the creation of
digital threads. It applies to product design, manufacturing, and quality inspection. It relies on
the XML standard and contains a library of XML schemas. It supports data integrity and
interoperability in implementing model-based enterprise and IoT [18]. QIF is used to structure
and report metrology information as parts are inspected by the DTL’s CMM.
5. Research and Development Efforts
Research within the DTL is conducted over a simplified product lifecycle (i.e., design,
manufacture, and inspection stages). Product design in the lab is set up to use Autodesk
Inventor and the designed product is manufactured using the equipment in the lab. Lastly,
inspection of the manufactured part is completed using the CMM. The following sections
describe the current ongoing research efforts that leverage the DTL.
Part Design to Support Prototype Demonstration
To fully leverage the work cell in the DTL, a part is designed to be cut on the CNC machine tool,
handled by the robots, and inspected by the CMM. This part is designed using Autodesk
Inventor. It is envisioned that future designs will be constructed using Inventor on the desktop
computer in the lab, and toolpaths will be generated using the same software. The part design
(shown in Fig. 6) is designed to be cut from wax stock.
Fig. 6. Part design to be cut on the CNC, handled by the robots, and inspected by the CMM
13
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 21
NIST AMS 100-68
August 2025
Equipment Digital Twins
This section discusses digital twins developed for individual devices and identifies specific
ongoing work. Individual device digital twins are under development for the collaborative robots
(see Section 2.3) and the CNC machine tool (see Section 2.4). Each of the device digital twins
are developed using the ISO 23247 standard and serve to test the standard as well.
5.2.1. Collaborative Robot Digital Twin
The DTL work cell includes two collaborative robots. As mentioned in Section 2.3, the robots
have six degrees of freedom and are designed for industrial manufacturing settings.
Collaborative robots, in the context of manufacturing, are primarily used for material handling
and machine tending. Universal Robots has an extensive list of data items that can be
collected through the UR-RTDE (Universal Robot Real Time Data Exchange) interface. This
interface allows third party developers to read real time data from the robots and send
commands to the robots. Using the UR-RTDE, MTConnect adapters collect each robot joint's
position, velocity, and acceleration for digital twin development. The orientation of the tool-
center-position (TCP), as well as data on the movement of the robot’s adaptive gripper
attachment, are also collected by the MTConnect adapter. All these data items are provided
by the built-in sensors installed by the vendor on the robots.
Figure 7 shows a deployed digital twin of one of our collaborative robots. To the left in the
figure is the physical robot looping through a basic drag-and-drop operation. On the top right is
the visualization of the robot’s motion based on simulated data produced by the model of the
robot (see Appendix C). On the bottom right is a visualization produced in MATLAB of the error
between the data from the simulated robot and the physical system.
Fig. 7. In progress robot digital twin for model validation and improvement
14
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 22
NIST AMS 100-68
August 2025
The digital twin model for the robot was developed in MATLAB Simulink is a simplified
mathematical description of the physical robot. Real-time data was integrated with the Simulink
model using MATLAB Simulink Desktop Real-Time, which provides a real-time kernel for
executing Simulink models, and the analytics application was developed in JAVA. Currently, this
digital twin does not have automated feedback to improve the model or affect the physical
system. This is one of the future research objectives of the DTL.
5.2.2. CNC Machine Tool Digital Twin
The DTL work cell includes a CNC machine tool. As mentioned in Section 2.2, the machine tool is
a five-axis desktop CNC. By deploying external sensors, researchers can gather more datasets,
such as vibrational and temperature measurements, and better inform digital twins for use
cases. The foundations, described as follows, for a data visualization digital twin are currently
implemented in the DTL. The CNC was modeled in Altair (see Figure 8), and the joint positions
were updated with data collected from the physical machine to reflect the CNC’s motion during
operation.
Fig. 8. Altair simulation of CNC machine tool
The model simulates the motion of the CNC by replaying historical datasets provided by the
CNC in the DTL. In the future, capabilities of Altair and other modeling tools to explore
multiple use cases of digital twins will be leveraged. Additionally, methods for streaming
collected data to update the digital twin in real time will be investigated.
Augmented Reality (AR) and Virtual Reality (VR)
While developing digital twins, an important topic cannot be avoided – the user interface.
Specifically, we are researching interface design for digital twins using AR/VR technologies,
which present significant challenges to manufacturers due to the unique expertise required.
One research effort in the DTL is developing an authoring tool [5] that addresses this challenge
by allowing manufacturing end users who do not have the programming skills to create AR/VR
15
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 23
NIST AMS 100-68
August 2025
interfaces for digital twins easily and quickly. The initial prototype demonstrates the ability to
parse MTConnect machine operational data and provide menu-based interfaces and a sandbox
of data visualizations to guide users through interpreting and analyzing the data.
6. Conclusion and Future Work
In this report, we outlined the vision and objectives of the DTL. Additionally, we described the
hardware and software setups for the DTL that enable it to support multiple research areas.
Some of these research areas result in the implementations of specific case studies that are
documented in this report. Lastly, we presented a list and description of standards that are
implemented in the DTL.
Looking forward, the lab will support the pursuit of various research objectives. First, we will
make data publicly available for external stakeholders. This will not only further establish the
DTL as a managed source of data but also increase the impact of research from the DTL. Future
work will address making the data available in a secure manner as well as allowing external
researchers to utilize both real time and historical datasets. Second, we will fully integrate the
CMM with the existing data pipeline. This would allow us to view data from the entirety of the
work cell in a standardized format and would make the process of digital twin development
much more efficient. Software for the CMM will need to be upgraded and we will collaborate
with the vendor to ensure that the CMM is compatible with the DTL’s data pipeline as the
pipeline evolves. Third, we will collaborate with external stakeholders to develop and test
methods for VVUQ (Verification, Validation, and Uncertainty Quantification) of digital twins.
The DTL will be instrumental in testing results of these collaborations. The research could
potentially impact standards for digital twins such as ISO 23247. Lastly, we will work with SDOs
to test and provide feedback on relevant standards. As a result, new use-cases could arise and
require different lab configurations, data pipelines, and enabling technologies.
16
PDF page 24
NIST AMS 100-68
August 2025
References
[1] Grieves, Michael. (2016). Origins of the Digital Twin
Concept. 10.13140/RG.2.2.26367.61609.
[2] U.S. National Economic Council. (2022). National strategy for advanced
manufacturing. The White House. https://www.whitehouse.gov/wp-
content/uploads/2022/10/National-
Strategy-for-Advanced-Manufacturing-2022.pdf
[3] ISO. 2020. ISO (DIS) 23247-1: Automation Systems and Integration - Digital Twin
Framework for Manufacturing - Part 2: Reference Architecture. ISO/TC
184/SC4/WG15
[4] Kibira, D., Shao, G., Venketesh, R., and Triebe, M. “Building a Digital Twin of a CNC
Machine Tool” Proceedings of the Winter Simulation Conference 2024.
[5] Simonson, A. and Shao, G. “An Authoring Tool for Manufacturing Digital Twin
Visualization” APMS 2024.
[6] Shao, G.; Frechette, S.; and Srini, V. An analysis of the new ISO 23247 series of
standards on digital twin framework for manufacturing, ASME MSEC 2023.
[7] Kibira, D. and Shao, G. “Data Requirements for Digital Twins of a Robot
Workcell,” Proceedings of the Winter Simulation Conference 2023.
[8] Kibira, D., Shao, G. and Weiss B. Building a Digital Twin for Robot Workcell Prognostics
and Health Management. Proceedings of the Winter Simulation Conference 2021.
[9] Kiran DR (2022) Chapter Twenty – Plan Location and Layout. Principles of Economics and
Management for Manufacturing Engineering, ed Kiran DR (Woburn, MA, Butterworth-
Heinemann), Chapter 20, pp 227-242
[10] Sakran, H. K., Mahbuba, H. M., & Jafer, A. S. (2016). A review of a basic concept of
cellular manufacturing. International Journal of Design and Manufacturing
Technology, 8(1), 30-37.
[11] Kibira, D., Shao, G., & Venketesh, R. (2023, May). Building A Digital Twin of AN
Automated Robot Work cell. 2023 Annual Modeling and Simulation Conference
(ANNSIM), (IEEE, Hamilton, ON, Canada), pp 196-207.
[12] Gaha, R., A. Durupt, and B. Eynard. (2021). Towards the implementation of the
Digital Twin in CMM inspection process: Opportunities, challenges, and proposals.
Procedia Manufacturing, 54:216-221.
[13] Helu M, Hedeberg T, (2020) Connecting, Deploying, and Using the Smart
Manufacturing Systems Test Bed. (National Institute of Standards and Technology,
Gaithersburg, MD), NIST Advanced Manufacturing Series (AMS) 200-2,
https://doi.org/10.6028/NIST.AMS.200-2
[14] MTConnect Institute (2023, Last Accessed September 2023) MTConnect Standard,
Version 2.2, Part 2 – Device Information Model, Standard. URL
https://docs.mtconnect.org/.
[15] ISO. 2021. 1SO (DIS) 23247-1: Automation Systems and Integration - Digital Twin
Framework for Manufacturing - Part 1: Overview and fundamental principles.
ISO/TC 184/SC4/WG15
17
PDF page 25
NIST AMS 100-68
August 2025
[16] ISO. 2021. ISO 10303-1: Industrial automation systems and integration - product
data representation and exchange - Part 1: Overview and fundamental principles.
ISO/TC 184/SC4
[17] MTConnect Institute (2023, Last Accessed September 2023) MTConnect Standard,
Version 2.2, Part 1 – Fundamentals, Standard. URL https://docs.mtconnect.org/.
QIF. 2020: The quality information framework standard. URL http://qifstandards.org/.
18
PDF page 26
NIST AMS 100-68
August 2025
Appendix A. MTConnect Device Configuration File
Below is the MTConnect Device Configuration file. The file is in XML format and lists the
available data items and associated metadata. This file covers data being streamed from all
machines in the laboratory. By using the Device Configuration file in combination with the
MTConnect standard documentation, stakeholders can understand what data attributes are
being collected and their context within manufacturing.
19
PDF page 27
NIST AMS 100-68
August 2025
20
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 28
NIST AMS 100-68
August 2025
21
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 29
NIST AMS 100-68
August 2025
22
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 30
NIST AMS 100-68
August 2025
23
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 31
NIST AMS 100-68
August 2025
24
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 32
NIST AMS 100-68
August 2025
25
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 33
NIST AMS 100-68
August 2025
26
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 34
NIST AMS 100-68
August 2025
27
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 35
NIST AMS 100-68
August 2025
28
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
PDF page 36
NIST AMS 100-68
August 2025
Appendix B. MTConnect Agent Configuration File
Below is the agent configuration file. This file configures the MTConnect Agent to connect to
the MTConnect adapters installed on each machine using their network addresses. The file also
sets the directory for the device configuration file, basic agent settings, and logging
configurations. Stakeholders can use the agent configuration file for similar setups and can
modify it to accommodate more machines and different data pipelines.
29
Figures on this page are available in the original PDF; text embedded in images has not been extracted.
来源：NIST · 制造研究 · nvlpubs.nist.gov