Artificial intelligence has made incredible strides in recent years, mastering skills such as vision, conversation, mathematical calculation, and even creativity. However, one critical ability has remained elusive: the capacity to measure or "feel" surfaces, a skill that typically requires a mechanical touch. But that’s all changing thanks to groundbreaking research from Stevens Institute of Technology.

Yong Meng Sua, a professor of physics at Stevens, explains the current limitations: "While AI has largely developed a sense of sight through advancements in computer vision, it still lacks a human-like sense of touch capable of distinguishing between different textures, like the roughness of newspaper versus the smoothness of magazine paper."

Now, researchers from the Center for Quantum Science and Engineering (CQSE) have turned this challenge into an innovative opportunity. Their recent study, published in the journal *Applied Optics*, details a cutting-edge method that empowers AI to feel surfaces through an intricate interplay of quantum technology and advanced algorithms.

The research team—including Director Yuping Huang, and doctoral candidates Daniel Tafone and Luke McEvoy—developed a unique laboratory setup incorporating a photon-emitting scanning laser alongside new AI models trained to interpret the surface qualities of objects. "This represents a significant fusion of AI and quantum technology," states Tafone.

The new system uses a specially designed beam of light that is pulsed at a surface to gather data. When this light reflects back, it carries a type of disturbance known as speckle noise, which is typically seen as an impediment to clear imaging. Surprisingly, this research team leveraged speckle noise to their advantage. By training their AI to understand these imperfections, they transformed them into valuable insights about the surface's topography.

The methodology involved probing 31 different types of industrial sandpapers with varying levels of roughness, ranging from 1 to 100 microns thick (for reference, a human hair is approximately 100 microns thick). Using advanced mode-locked lasers, they acquired light pulses that targeted the surfaces, then analyzed the reflected data through their learning model.

Initial tests revealed a root-mean-square error (RMSE) of about 8 microns, but with iterative refinement and additional sample processing, the accuracy improved to within an impressive 4 microns—comparable to the finest industrial profilometers available today. Notably, the system excelled with smoother surfaces, such as diamond lapping film and aluminum oxide.

This revolutionary technology has immense potential across several fields. For instance, in the realm of dermatology, it could revolutionize skin cancer detection, where subtle differences between benign moles and malignant melanomas often lead to misdiagnosis by human examiners. "Minor variations in mole roughness, undetectable to the naked eye but measurable with our quantum-based system, could be crucial for differentiation," explains Huang.

Quality control in manufacturing is another critical application, where accuracy at an extremely small scale can mean the difference between creating a flawless component and producing a defective part that could lead to serious mechanical failures.

As automated systems, such as LiDAR technology in self-driving cars and robotics, become increasingly incorporated in our daily lives, this novel method could significantly enhance their capabilities by allowing for precise measurement of surface properties at incredibly small scales.

In a future where machines could not only see but also feel, the possibilities for innovation are endless! Stay tuned for more exciting breakthroughs in the world of AI and quantum technology!