“`html

Addressing precompetitive challenges, such as facilitating interaction among digital twins, U-M and ASU are searching for industry collaborators for a prospective NSF research center

Striving to break down obstacles that hinder digital twins from fulfilling their potential to enhance manufacturing, the University of Michigan and Arizona State University are inviting industry collaborators to engage in a new Center for Digital Twins in Manufacturing.

“Everyone is developing digital twins, but we’re focusing on creating the connections that allow digital twins to collaborate—to be modular, reusable, and sustainable,” commented Dawn Tilbury, the Ronald D. and Regina C. McNeil Department Chair of Robotics at U-M.

“Teams within this center will tackle precompetitive challenges that can offer widespread advantages to all who create and utilize digital twins, aiding in the expansion of these remarkable technologies to enhance manufacturing performance, quality, and operational uptime.”

With sufficient industry stakeholders, the team aspires to officially launch the Industry-University Cooperative Research Center, financed by the National Science Foundation. For such centers, which are situated in existing laboratories, the NSF allocates $1.5 million over five years to cover administrative expenses, while industry partners contribute annual fees of $90,000 to support the research initiatives.

An informational session for interested industry representatives is scheduled for 1-4 p.m. on July 23. Participants can register online.

The session will feature presentations about the center, its objectives and operations, and the advantages to members. Breakout and discussion segments will also be included to uncover the most pressing topics for potential members and identify specific industry challenges.

“This gathering will serve as an excellent opportunity for prospective industry partners to gain deeper insights into the center’s vision, comprehend the benefits of membership, and influence the initial wave of projects,” stated Wenlong Zhang, ASU associate professor of robotics and manufacturing systems.

A digital twin represents a virtual model of a device that communicates with the actual device, continuously updating itself to mirror the current state of the device it is simulating. For example, a digital twin of a milling machine may illustrate the progress of a component carved from a metal block, checking in every tenth of a second to ensure its model syncs with the real machine. The twin can make forecasts regarding part quality, scheduled maintenance, and more. However, the absence of standardization complicates the implementation of digital twins in manufacturing.

Currently, Tilbury notes that the majority of digital twins in manufacturing are designed for individual devices. If a factory possesses a digital twin for an entire system or production line, it is likely a customized software solution that needs to be rewritten when any part of the system is modified—for example, if a machine undergoes an upgrade. Meanwhile, numerous organizations are offering digital twin software for single machines. Connecting them with the twins of preceding and succeeding machines on the line would facilitate enhanced coordination both upstream and downstream.

The proposed center intends to tackle various challenges associated with digital twins in manufacturing from multiple angles. They aim to develop generalized digital twins for specific machine types—for instance, a 3D printer that can be adapted to represent a particular machine. These twins are designed to be reusable, extendable, and maintainable, allowing for system expansions, accommodating natural machine degradation, and showcasing improvements post-repair or replacement of parts.

Additionally, the center’s objectives include:

• Measuring and minimizing uncertainty in digital twins
  • Enhancing predictions and assisting factory managers in evaluating their accuracy
• Creating digital twins for human-robot collaboration
  • Digital twins of human counterparts, monitoring only the elements of the human that are pertinent to the task, could assist robots in becoming better teammates while alleviating privacy concerns
• Utilizing digital twin software for simulation and “what-if” analyses
  • In simulation mode, independent of a physical machine, digital twin software could empower factory managers to train staff and explore potential machine upgrades or reconfigured factory layouts
• Autotwin: Software capable of generating and operating digital twins

To evaluate concepts developed through the center and ensure they function as intended, U-M provides the SMART 4.0 testbed, featuring a “connected factory” of mobile robots, computer-controlled machines, and 3D printers interconnected via open process automation. ASU contributes its own connected smart manufacturing system, encompassing robotics, programmable logic controllers, smart sensors, and RFID tracking for milling and additive manufacturing.

Tilbury is also the Herrick Professor of Engineering and a professor in mechanical engineering and electrical engineering and computer science.

“`