Researchers from the National University of Singapore (NUS) have come up with a new way to counter the DeepKey name. Developed in just eight months, this security innovation uses two-dimensional (2D) -material tags and artificial-intelligence (AI) authentication software.
Compared to conventional anti-blocking technologies, DeepKey works faster, achieves highly accurate results, and uses stable identification tags that are not easily damaged by environmental conditions such as extreme temperatures, chemical spills, UV exposure, and humidity. This new testing technology can be applied to a variety of high-value products, from drugs, jewelry, and electronics. For example, DeepKey is suitable for the tagging of COVID-19 vaccines to enable fast and reliable testing, as some of these vaccines need to be stored at an extremely cold temperature of -70 ° C.
Led by Asst Prof Chen Po-Yen and Asst Professor Wang Xiaonan from the Department of Chemical and Biomolecular Engineering at the NUS Faculty of Engineering, the team’s 2D material security tags display physically inoperable patterns of action (PUF patterns ), which are randomly generated by a system. crimping the 2D material thin films. The complex patterns of 2D material with multi-scale features can then be sorted and tested with a well-trained in-depth learning model, enabling reliable proofing (100 percent accurate) in less than 3.5 minutes.
There are usually several anti-counterfeiting technologies that use PUF patterns, including complex manufacturing, special and tedious reading process, long trial period, adequate environmental stability, as well as being expensive to do.
“With this research, we have addressed a number of bottles that come in other ways,” said Professor Asst Wang. “Our 2D-material PUF tags are environmentally stable, easy to read, simple and inexpensive to manufacture. In particular, the adoption of in-depth learning greatly accelerated the overall validation, pushing our invention one step further to practical use. “
The researchers published their findings in a scientific journal Issue on December 2, 2020. This study was conducted in collaboration with researchers from Anhui University of Technology and Nanyang University of Technology.
Stable, simple and scalable process to create PUF tags
Surprisingly, the researchers do not need any special equipment to create the secure tags. They can be simply made with a balloon, a bottle of 2D material dispersion, and a brush.
“First, we inflate the balloon and brush over its surface with 2D-viscous ink. After air dries overnight, we disinfect the Due to the mechanical imbalance between the 2D material and the latex substrate, large area PUF patterns will be generated during the shortening.These PUF patterns can be cut to the required size thereafter, and usually hundreds of them can be done at one time, “said Dr Jing Lin, a member of the research team.
Next, the researchers take a snapshot of the PUF tag with an electronic microscope, which is then synchronized to their innovative software to go through the deep learning “sorting and validation” process. . “The whole process takes less than 3.5 minutes, and most of it is spent waiting for the reading from the electronic microscope. The proof itself is very fast, in less than 20 seconds,” he explained. and Dr. Jing.
Quick validation using AI deep learning algorithms
All PUF-based technologies have ultra-high coding capabilities due to the large number of unique patterns that can be theoretically realized. However, the high coding capability also leads to a long validation period, as “find and compare” pattern verification is required within a large database. This trade-off between high coding capability and long validation time often limits PUF-based anti-counterfeiting tags from practical applications.
“With our new technology, we are breaking this enduring trade-off between high coding capability and long test time by using 2D material that can be sorted and deep learning algorithms,” said Asst Prof Wang.
Initially, the researchers used different 2D materials to make PUF tags with recognizable AI features. Second, they trained in an in-depth learning model to make a two-step verification apparatus. “We used the in-depth learning model to pre-sort the PUF patterns into subgroups, so the search-and-compare algorithm is performed in a much smaller database, which shortens the overall validation time, “explained Asst Prof Wang.
Currently, polymer wrinkle-based tags are the only available technologies similar to this NUS innovation. Wrinkled polymer tags are proven based on the surface patterns just like the novel 2D-material tags. However, their current certification requires one-on-one drag and match feature, which is slow and shows only 80 percent reliability. NUS team validation is driven by deep learning, so it is much faster, and reaches nearly 100 percent authentication accuracy.
Moreover, in contrast to the wet chemistry preparation of polymer wrinkle-based tags, which involve the use of harmful organic chemicals and UV light, the manufacturing invention of NUS researchers is much faster and safer.
The next steps
The NUS team has filed a patent for their invention and now intends to push this technology one step further. “We are looking for better, faster and more robust read and validation methods for our PUF tags,” said Asst Prof Wang.
The team has already begun researching alternative reading methods to further shorten the processing time. “Furthermore, such information naturally encoded by the PUF tags could be made more secure by being kept on a blockchain, so that the supply chain goes to and strict quality control, “added Asst Prof Wang.
###
Disclaimer: AAAS and EurekAlert! they are not responsible for the accuracy of press releases posted to EurekAlert! by sending institutions or for using any information through the EurekAlert system.