Mark Zuckerberg’s vision for AI avatars, shared during his recent discussion with NVIDIA founder Jensen Huang, has generated a bunch of articles tying the concept to digital twin technology.
Zuckerberg wants to help creators and small businesses to create AI replicas of themselves for public interactions.
His company Meta just launched its AI Studio, where users can create digital versions of themselves to interact with followers.
How much does Zuckerberg’s vision align with digital twin technology as defined today?
Avatars vs. Digital Twins
Today’s AI avatars do basic tasks like responding to comments and direct messages. Media commonly refers to the AI representations as digital twins which is misleading.
Current avatars are basic digital representations of a person’s general persona and style. They rely on user sourced, static or at best semi-dynamic data.
With limited data integration, today’s avatars come nowhere close to the technical definition of digital twin. But in the future, the technology is bound to incorporate more data for more personalized and context-aware interactions.
A virtual replica of a person would become a digital twin when it incorporates real-time biometric, location, and activity data. The twin would mimic its real-world counterpart’s responses and decision-making tendencies. At this point, the digital twins could reliably fill in during virtual meetings and provide proactive suggestions in anticipation of future exchanges.
So for now AI avatar is more accurate.
More Digital Twin Updates
Now let’s get into the state of digital twin technology of today. Subscribe to my YouTube channel to stay up to date as digital twins surge in popularity and use the comment section to suggest updates for future roundups.
ASICS Personalization Studio
ASICS is experimenting with on-demand sneaker personalization using advanced digital twin and 3D printing technology at a new microfactory in Paris.
At the ASICS Personalization Studio, a detailed scan of the runner’s foot captures its exact shape and dimensions.
A digital twin of the foot is then created using the Dassault Systèmes 3DEXPERIENCE platform.
The sockliner design is generated based on the detailed digital model, ensuring a perfect fit. The custom sockliner is then 3D printed with tailored cushioning and support.
A custom fit improves alignment and pressure distribution, boosting performance and recovery.
Dassault and ASICS plan to test the technology in Japan in 2025.
In a press release, ASICS President and COO Mitsuyuki Tominaga said: “Our partnership leverages the advanced technology and expertise of two industry-leading companies to provide optimal value to each customer, helping them live healthier, happier lives with products that respond to their individual needs.”
LuxCreo’s 3D Printing Smart Factory is part of the personalization studio. The Chicago-based company’s special resin is stretchy and keeps its shape. The material is also used for LuxCreo’s custom dental and medical products.
GAO Urges DoD to Adopt Digital Twins for Hypersonic Weapons
The US lags behind Russia and China in hypersonic weapons because it isn’t fully using modern engineering tools like digital twins.
Despite spending billions, the Pentagon hasn’t fielded any hypersonic weapons, which fly more than five times the speed of sound.
Cost estimates are tricky because the Pentagon lacks experience in the technology. So officials rely on expert opinions, which can be biased.
In a new report, the Government Accountability Office says four out of six hypersonic weapon programs are not using modern digital engineering tools or asking for user feedback. They are:
- Army Long-Range Hypersonic Weapon
- Navy Standard Missile 6 Block IB
- Air Force Hypersonic Attack Cruise Missile
- Air Force Air-launched Rapid Response Weapon
The budget office says modernizing their engineering tools would accelerate development, reduce costs, and lead to more effective and user-friendly hypersonic weapons.
The defense department has been pressuring its programs to digitally transform defense systems. In late 2023, the defense department issued its digital engineering directive, DoDI 5000.97, which requires detailed digital models as the primary method of communication in new programs unless exempted.
NVIDIA Tools for Accelerating Humanoid Robots
NVIDIA has released new tools aimed at accelerating the deployment of humanoid robots worldwide.
The tech giant just announced a suite of tools for robot simulation, learning, and training. A new humanoid robot developer program provides early access to the tools, with companies like Boston Dynamics and ByteDance already participating.
NVIDIA says its NIM microservices can reduce robot deployment times from weeks to minutes. The MimicGen NIM microservice creates synthetic motion data, while the RoboCasa NIM microservice makes simulation-ready environments. Osmo, NVIDIA’s cloud-based service, simplifies robot training and simulation. The company says it cuts development times from months to under a week.
Teleoperation, which captures human actions to train robots, can be done faster and cheaper with NVIDIA’s AI and Omniverse digital twin tools.
The software is compatible with the Apple Vision Pro, which captures human demonstrations that are simulated and expanded into large datasets for training robots.
NVIDIA has been vocal about humanoid robots, investing heavily in their development and capabilities. Program members get early access to foundational models in NVIDIA’s Project Group. The project, announced earlier in 2024, uses NVIDIA’s tech to teach robots how to move and interpret human language.
Skild AI Funding for Foundational Modeling
A robotic startup could use a foundational model with their hardware as a starting point. Since the model already knows a lot, the startup can focus on teaching the robot to perform tasks that can’t yet be automated.
Skilled AI has secured $300 million to build a foundational AI model to power robots for real-world applications. The round of Series A funding brings the San Francisco-based startup’s valuation to $1.5 billion.
The company says the funding will go toward advancing AI to make robots more adaptable and useful across industries. by Lightspeed Venture Partners, investors include Coatue, Softbank, Jeff Bezos, Sequoia, Felicis Ventures, and Amazon.
Skild AI says its foundational model is trained on a significantly larger and more diverse set of data compared to competitors. The company says it’s working on hiring experts in AI, robotics, engineering, operations, and security to help bring advanced AI into physical reality.
El Dorado Refinery Digital Twin
Delek’s El Dorado refinery is now using leading-edge digital twin technology to better run the facility that processes 80,000 barrels of crude oil a day.
A virtual model of the Arkansas facility reflects the petroleum refinery’s state in real time. The digital twin provides 3D views with data from controls, maintenance, inspections, and sensors.
The refinery uses Mobideo’s digital twin platform powered by R2S technology to create and manage the models. Delek US Holdings says the technology makes it easier to plan and understand complex data.
The company says virtual equipment checks save time and money. They also credit the digital twin with improving teamwork and resource management.
The Mobideo Digital Twin platform is geared toward Shutdown, Turnaround, and Outage (STO) planning.
Delek’s refineries processes more than 300,000 barrels of crude oil daily.
Softbank Predictive Accident Prevention
Your visibility is under 50 meters as you drive on a curvy road on a foggy night. You have about 3.6 seconds to react when you finally see the broken-down truck blocking your path.
The tragedy to follow can be prevented using state-of-the-art digital twin technology. That’s according to SoftBank, which details its predictive accident prevention technology in a new report called Smart Roads Ahead.
The Japanese telecommunications company says its experiments confirm the technology can make roads safer.
Digital twins merge real-time data from connected cars and road sensors. The combined data creates a virtual model of the road environment. The digital twins use state estimation and sensor fusion to track vehicle positions accurately.
When a potential risk is detected, the system alerts the would-be victim to avert tragedy.
Softbank says the technology works for non-connected vehicles using data from roadside sensors. The unconnected driver would get the alert on their infotainment system or smartphone.
Predictive Cattle Breeding
Digital twins are now used to pair cattle for breeding. A major Australian cattle producer uses the genetic predictive technology to prevent inbreeding and predict breeding outcomes.
The Australian Agricultural Company detailed the technology to Beef Central. The tool extracts info from DNA samples taken from cattle.
The data is analyzed against breeding records, animal performance metrics, and databases containing genetic info from many cattle. The combined data goes into digital twins that simulate and predict outcomes. The detailed digital models provide insights to improve overall herd quality and productivity.
Continuous updates are made to the virtual replicas with new genetic data as it’s updated.
The Australian company now plans to integrate the technology more broadly across its breeding programs.
Reconcile Reality through Simulation
Robot success increases by over 67% using digital twins of real-world locations for training.
That’s according to researchers from the US and Germany who developed a new method for training robots with computer simulations called real-to-sim-to-real or RialTo.
- First, the physical location is scanned to create a detailed digital copy.
- With that, a digital twin is built that reflects the physical space in real-time.
- The robot software operates in the virtual environment, letting it see and interact with objects as if they were real. Robots practice tasks in the digital twin, safely learning and improving their skills virtually.
- After training in the simulation, the research team says the robots perform the tasks in the real world with much greater success.
The study used Franka Emika Panda robotic arms. Tasks performed included stacking dishes and placing books on a shelf. The researchers are affiliated with the Massachusetts Institute of Technology, the University of Washington, and the Technical University of Darmstadt. The study was partly funded by the Sony Research Award Program, the US federal government, and Hyundai.
Edge Cache Optimization for Wireless Networks
A new digital twin method makes wireless networks faster and more reliable by optimizing edge caching, according to computer scientists from North Carolina State University.
The researchers developed the new method called D-REC, which uses a virtual copy of the network to predict what data users need. It uses reinforcement learning to make better decisions about data storage.
According to the researchers, D-REC prevents network overloads by predicting and adjusting data storage.
The research was published in the IEEE Journal on Selected Areas in Communications.
Stay Updated
Remember to subscribe to my YouTube channel for deep dives on the past, present, and future of digital twin technology.