Cloud Hosting

techniques Photo courtesy of Zheng Feng

Opinions expressed byDigital Journalcontributors are their own.

According to a report on the Chinese-American Railway Transportation Joint Research Centers website, the project Wheel-Track Wear Indicators, Calculation and Testing Methods for Heavy-Haul Freight Cars has been successfully completed and passed the final review by CRRC officers and professors from the engineering department at the University of Illinois Urbana-Champaign. This project was led by Zheng Feng, who designed a user-friendly and easily transportable software testing platform to assist with fast and convenient simulation studies. This platform allows companies and research institutes to use computer simulation to calculate wheel-rail wear on running heavy-haul trains, predict wheel life, and determine railway maintenance cycles.

In 2015, Zheng turned down several universities that offered him admission and went through another year of preparing for entrance exams and completing additional research to gain admission to an American university with a top engineering program: the University of Illinois Urbana-Champaign, where he pursued a graduate degree in mechanical engineering.

During his initial years at the University of Illinois, Zheng primarily worked on courses that focused on manufacturing and control algorithms to be used in traditional industry areas while concurrently developing his skills in information technology. With an impressive first-year academic record, he began keeping his hands busy on research in the lab. His research served to enhance what he was learning in the classroom, assuring him that he was headed down the right path.

Through projects like developing the wheel-track wear calculations software platform, Zheng has helped to lay the groundwork for future studies to utilize computer science techniques, such as writing calculation and simulation software platforms, to address challenges in traditional industries.

When it came to Zhengs second project, he participated in a research project through his advisors recommendation, which became the step he needed. With a background in software development and machine learning technology and performance in wheel-track wear calculation and prediction platform development, he was trusted as an assistant in multi-task Gaussian process learning on ultrasonic welding data.

Ultrasonic welding is different from the traditional welding method. It provides better welding quality and a cleaner surface near the welding spot. This technology has been widely used for joint batteries used in electric cars. However, due to the complicated physical process during welding, it is often hard to identify the quality through traditional methods like mechanics analysis and thermal identification.

This was a challenge for Zheng at the time. Zheng spent a lot of time using different techniques to connect the welding performance with the online signals through thermal and material features generated from the microstructures. Then, unfortunately, he failed. However, an idea was sparked: how about using untraditional approaches like machine learning to figure it out?

By combining welding parameters and online signals with the final welding quality, he thought he should be able to perform pattern recognition and feature generations to control the welding process and improve the final welding quality. Despite initial setbacks, his approach yielded the results he was hoping for, even surpassing expectations and serving to enhance ultrasonic welding quality through optimal control algorithms.

Armed with another masters degree in computer science, Zheng Feng has become a software developer and cloud infrastructure researcher at VMware by Broadcom, a well-known name in high-tech innovation. Zheng remains committed to moving the field of engineering forward.

Read this article:

Zheng Feng: Addressing challenges in traditional industries with computer science technology - Digital Journal