Dr. Cao’s Journey: From Hanoi to Tokyo in the Pursuit of Technological Mastery
From Hanoi to Tokyo, Dr. Cao — a Vietnamese AI expert — has turned his aspiration to “understand and master technology” into reality in Japan. With a strong research background in machine learning and image processing, he leads MITECH in delivering AI-based visual inspection solutions for Japanese manufacturers, tackling challenges related to labor shortages and quality control.
His journey represents a fusion of Vietnam’s creativity and Japan’s spirit of monozukuri—precision, craftsmanship, and manufacturing excellence.
English Translation from: https://tama-innovation-ecosystem.jp/tama-area/interview/5238
As part of this initiative, small and medium-sized enterprises, startups, large companies, and universities both within and outside the region are collaborating on projects aimed at solving local issues, as well as establishing member-based organizations (communities) that enable interaction among diverse stakeholders. These efforts are designed to promote innovation. Through this interview series, we aim to highlight leading companies driving innovation in the Tama area, and to share the region’s unique strengths with our audience.
MITECH Co., Ltd. is a startup that tackles challenges in manufacturing by leveraging AI and cutting-edge image processing technologies. The company’s founder and CEO, Thang Cao, is a Vietnamese-born Ph.D. in engineering. Why did he decide to start a business in Japan, and how has he approached technology throughout his career? We spoke with Mr. Cao to learn more about MITECH’s strengths and its future vision.
Came to Japan at the age of 32 on a scholarship
Interviewer: We understand you studied at Hanoi University of Science and Technology. What was your field of study?
Cao: I majored in electronic engineering, focusing on semiconductors and circuit design. When I graduated, Vietnam was not yet as developed as it is today, and there were limited opportunities for challenging new fields such as research. I originally wanted to pursue research, but due to the lack of options, I joined an IT company in Hanoi and worked there for nine years.
Interviewer: Why did you decide to study and work in Japan?
Cao: During my nine years of working in Hanoi, Vietnam grew little by little, but the country still had not advanced into core technology fields. We mainly sold computers and software from foreign manufacturers. The internet was not widespread yet, and dial-up was still the norm. I had been preparing to study in the U.S., so I quit my job to improve my English and apply for scholarships. At that time, all I really knew about Japan was “Made in Japan.”
While waiting for the results of the U.S. scholarship, I happened to read about a Japanese scholarship in a newspaper and applied. I received the acceptance from Japan before hearing back from the U.S., which ultimately led me to Japan—at the age of 32. Not so young anymore (laughs).
I started studying Japanese from scratch and entered Ritsumeikan University on a scholarship. Fascinated by machine learning, I began researching that field in 2003—what we now call AI. I completed my master’s and doctoral studies, earning my Ph.D. in March 2008. Afterwards, I joined a company in Shiga Prefecture because I wanted to understand why Japan had developed so much, how Japanese people work, and what cultural values shape Japan’s work ethic. I wanted to see with my own eyes why Japan could create technologies and products that reach the world. At the company, I worked on developing inspection equipment, robot control software, and an automatic clothing-folding machine.
Passion for work—the source of Japan’s manufacturing strength
Interviewer: After working in Japan, did you find the answer you were seeking?
Cao: Yes, I did. What struck me most was the passion Japanese people have for their work. Sometimes we worked from 8 a.m. to 10 p.m., but it wasn’t just long hours—the quality of the work time was incredibly high. People always took notes, handled tasks carefully, and maintained organized workspaces. Their seriousness toward every task left a strong impression on me.
Another striking aspect was their dedication to learning. Whenever a new technology or patent appeared globally, the team would immediately hold study sessions and discuss differences from existing technologies and potential applications, then incorporate improvements into actual products. Through this attitude, I came to understand part of the strength behind Japanese manufacturing.
Interviewer: What motivated you to start your own company in Japan?
Cao: I always wanted to continue research, so after leaving the company in Shiga, I worked at the University of Electro-Communications and the University of Tokyo on supercomputer-related projects. The work was rewarding, but my desire to pursue AI and machine learning grew stronger. So I made a bold decision to leave the University of Tokyo and establish my own company specializing in AI and image processing. That became MITECH, founded in 2018.
Strengths rooted in deep expertise in AI
Interviewer: Could you tell us more about your company’s business?
Cao: We develop visual inspection software for the manufacturing industry using AI and image processing. The software automatically checks products on factory lines using cameras to detect defects—for example, a loose cap, a torn label, scratches, or foreign matter.
Traditionally, these inspections were done by human workers, but labor shortages have made it increasingly difficult to secure inspectors. As a result, the demand for AI-based automation has risen dramatically. Recently, I feel the technology has finally reached a level suitable for real-world use.
Image-based inspection has existed for a long time, but older techniques struggled with many cases. With AI, judgments that were previously difficult for machines—yet obvious to humans—are now becoming possible. Advances in CPUs, GPUs, and hardware performance have played a major role in this shift.
I have researched AI since my university days, so I understand the algorithms deeply. This allows me to build systems from scratch and respond flexibly to requests for improved accuracy. When issues occur, I can adjust the internal logic. This deep understanding is a key strength that differentiates us from other AI companies.
AI brings unprecedented advances to image processing
Interviewer: What advantages does AI-based image processing have over traditional methods?
Cao: For example, humans can instantly tell the difference between a perfect circle and a slightly distorted one. But traditional image processing struggles with such subtle differences. It relies on human-selected parameters like aspect ratio, area, or brightness, and becomes ineffective when shapes get complex. Typically, traditional systems use around 10–20 parameters.
AI is completely different. Instead of 10 parameters, it can automatically analyze tens of thousands, millions, or even billions to determine differences. In short, AI can translate “something feels different” into numerical reasoning. That’s a breakthrough that was previously impossible.
From semiconductor wafer inspection to predicting sleep apnea
Interviewer: What kinds of products or technologies have you developed?
Cao: Semiconductor wafer inspection: detecting minute dents, particles, or scratches before processing.
Medicine can inspection: checking for dents, scratches, foreign matter, or even insects.
Smartphone cover glass inspection: detecting defects in real time as each glass passes every 2–3 seconds on the line.
Sleep apnea prediction: predicting 30 seconds before an apnea event so devices can intervene and prevent danger.
We also work with semiconductor inspection equipment manufacturers in the Tama region, which aligns with our vision of contributing to the local economy while growing together with local companies. Close collaboration allows for quicker responses and innovation, benefiting both the region and our business.
Leveraging talented young engineers from Vietnam
Interviewer: What does your current development structure look like?
Cao: We have five engineers in Japan and fifteen in Vietnam. Vietnamese engineers excel at logic design and software development. Meanwhile, due to Japan’s aging population, hiring engineers is becoming increasingly difficult. So we train Vietnamese engineers in Japanese and support their opportunities to work in Japan. I want to build a relationship that benefits both countries.
We also have strong participation from students and graduates of Hanoi University of Science and Technology, my alma mater. By creating an environment where young talent can gather, we hope to continue strengthening both our human resources and our technology. Being located in the UEC Alliance Center, we also plan to offer internships and part-time opportunities to students at the University of Electro-Communications.
Strong collaboration potential with automated machine manufacturers
Interviewer: What types of companies would you like to collaborate with?
Cao: Primarily, automated machine manufacturers—companies that build equipment to transport products on production lines or position them for inspection. These companies are usually strong in mechanical engineering but have less expertise in image processing or AI. That’s where we can complement them.
We also have strengths in data analysis and IoT, so collaboration in those areas is possible as well. By combining our technologies to create “products that sell,” we can expand business opportunities. We are currently exploring ways to reach end users directly, but joint product development with other companies is also a promising direction.
Building more technically valuable projects
Interviewer: Finally, what are your goals for the future?
Cao: We hope to steadily grow our company while actively incorporating new technologies. Working with talented engineers to contribute to society through technology is our mission. Of course, increasing revenue is welcome, but rather than chasing numbers, we want to build a portfolio of technically valuable projects. Instead of rapid expansion, we aim for steady growth, creating a hub for advanced technology and giving back to society. That is the future we envision.