Relying upon the speed and efficiency of computers has become so second nature that most people don’t stop to think about what drives it all—unless you want to see things move even faster. And Kevin Gu does.
When Gu came to the U.S. over twenty years ago to pursue a PhD, computer speeds were already getting faster. “As computers ran at higher and higher speeds, so-called “signal and power integrity” issues became a major challenge,” Gu described. “Engineers needed better tools to simulate, model and analyze the effects caused by the higher operating frequencies, and to test and validate new design approaches.”
The frequency of a computer processor is measured in gigahertz (GHz), which indicates how many billions of cycles are completed in one second—one hertz means one cycle per second. Gu was drawn to working in the less-explored sub-terahertz (sub-THz) arena; this is the higher end of cycles, between 90 and 300 GHz. Gu’s company, Astrabeam, seeks to leverage the latest developments in sub-THz to improve the capacity of remote-sensing and radar-sensing devices, like satellites and robots.
“Going above 100 GHz is really entering a new frontier,” said Gu. “It enables greater potential for communications and radar sensing. There’s lots of spectrum available there. It’s like the Wild West—we are still figuring out how to make it work. But more open, untapped bandwidth has great potential. For example, when you go to higher frequencies, you can design more compact devices that may be easier to deploy in remote situations.”
Gu pursued his PhD at the University of Washington, studying computational electromagnetics. “This was a growing area back then,” said Gu. “I was specifically working on connecting and running computers at higher speeds.” Gu’s work explored a numerical approach to solving Maxwell equations, which describe how electric and magnetic fields relate to each other.
After completing his PhD, Gu headed east to join the IBM T. J. Watson Research Center in Westchester County, outside New York City. He worked on a range of projects involving millimeter wave—the range of electromagnetic frequencies between microwaves and infrared. “As time went on, IBM was primarily focused on creating IP and licensing to other big companies,” he said. “But I could see that the technology was rapidly evolving and there were more opportunities in higher frequencies.”
“One of the fundamentals of chip technology is that whatever makes it work keeps it scaling,” Gu added. “That led me to a new idea—to push for a lower cost device that would push frequencies and speeds even higher.”
Gu described how some of his colleagues and peers wanted to teach or do research, but he was more interested in commercializing technology. “Some of us had this common goal, to push the tech to the next level, and that could lead to commercialization opportunities. Eventually, we thought, why not collaborate and use some NSF grants to try it out?” That led to the creation of Astrabeam.
Gu emphasized the important role that the NYC Innovation Hot Spot had played throughout his startup journey. He initially learned about the Hot Spot from doing Google searches for startup resources in New York. “I found out about the summer 2022 regional NYC I-Corps training,” said Gu. “I had already been working on Astrabeam, and my Phase 1 NSF STTR had just been accepted. I wasn’t completely sure about the market fit or who the customer would be. I thought it would be good to get a better understanding of the business and commercialization side of things.”
“It really helped to step away from the technology and do all of those interviews,” Gu added. “Key to making a successful startup is figuring out the market fit, and we got that from I-Corps. We had to understand who wants to use the technology, rather than just focusing on the development of the technology itself. It was a really good lesson to learn!” Gu soon after participated in the SBIR National I-Corps, bringing the company to another 100 interviews in Spring 2023. “There was so much more we needed to do to understand customer needs,” Gu said.
With his company, Gu hopes to offer a range of technologies at a higher frequency than is currently available. Going to higher and higher frequencies, believes Gu, will offer new applications of better sensing accuracy while demanding less weight, and at a smaller size, which could have many benefits to different sectors.
Gu envisions numerous uses for sub-THz. For example, his radar sensors can detect human vital signs to improve automotive in-cabin safety. They can also support level measurements inside a tank (whether holding oil, water, or other substances). In the industrial automation arena, the standard is to use robots outfitted with optics-based sensors—relying on cameras and light; Gu’s radar sensors would be able to overcome environmental conditions such as dust or smoke.
Astrabeam benefited from several other NYC Hot Spot services post I-Corps, including two SBIR/STTR Proposal Writing Assistance workshops. One of those sessions enabled Gu to receive several hours of consulting hours of support that led to a successful phase 1 NASA SBIR contract in 2023. The second workshop was about writing proposals to the US Department of Defense. Gu also shared that the Storytelling for Scientists workshop held in April 2024 helped him to improve his presentations in pitch competitions.
Gu addressed the many challenges he’s faced on the road to commercialization.
“For me, a big piece of the learning has been centered on customer discovery,” Gu added. “It’s a significant thing for a tech person like me. I keep reminding myself that the business side is equally important—understanding the market dynamics, finding the customer, and developing something that they want to use. That is something I still need to improve upon and strengthen as a founder—it’s critical to our early-stage development.”
After bootstrapping Astrabeam, Gu is accelerating its R&D and commercialization with support from Columbia MBA students, UIUC engineering PhDs, robotics enthusiasts from the local New York Robotics Network, and partners at Pennsylvania State University and Advanced Science and Research Center (ASRC) of the City University of New York. Astrabeam is currently participating in a Westchester County, NY-based accelerator program, Element 46, that runs through fall 2024, and is one of ten semiconductor startups participating in a fall accelerator program from Plug and Play NeoCity in Osceloa County, FL. Astrabeam has also joined two CHIPS and Science Act-funded Department of Defense microelectronics hubs.
Gu is also beginning to look at private opportunities for funding, including angel and venture capital pre-seed funding, mostly focused on radar sensors with autonomous robotics applications. “We did 200 customer discovery interviews just for the robotics space,” he added. “The benefits and insights I’ve gained from the NYC Hot Spot continue to impact our work.”
コメント