How Digital Twins Could Protect Manufacturers from Cyberattacks
NIST News, February 23, 2023
At the U.S. National Institute of Standards and Technology and the University of Michigan, researchers have combined digital twin technology, machine learning, and human expertise into a cybersecurity framework for manufacturers. The researchers constructed a digital twin to mimic a three-dimensional (3D)-printing process, supplemented with information from a real 3D printer. Pattern-recognizing models monitored and analyzed continuous data streams computed by the digital twin as the printer created a part, then the researchers introduced various anomalies. The programs handed each detected irregularity to another computer model to check against known issues, for classification as expected anomalies or potential cyberthreats; a human expert made the final determination. The team found the framework could correctly differentiate cyberattacks from normal anomalies.
No comments:
Post a Comment