More than 100 students from around the globe worked to solve real-world challenges with artificial intelligence (AI) technology during the 2021-2022 AI Cup, an annual competition hosted by TE Connectivity, a world leader in connectivity and sensors.
Twenty-three teams of students across three continents developed innovative solutions supported by AI technology to solve the challenges at the manufacturing sites. Participating teams were encouraged to visit TE factories at the beginning of the competition to understand the challenges of the site and their ideas were verified and adjusted through either on-site or remote testing.
The winning solution emerged from a team from Universidad de Sonora in Mexico. These students developed an AI vision inspection system with a novel synthetic defect picture generator for TE’s automotive plant in Hermosillo, Mexico. This solution is designed to solve the challenge that product lines often rely on manual inspections and cannot introduce AI visual inspection system due to the insufficient sample size of defects. The tool was nearly 100% effective in detecting defective parts during the testing.
"TE AI Cup is one of the platforms that TE established to promote industry-university collaboration and cultivate the next generation of engineers. This year, I am very proud and pleased to see more teams around the world joining the competition," said Dr. Roberto Lu, the founder of TE AI Cup and vice president technology, TE Global Operations. "As a global industrial technology leader, TE strives to provide a platform where engineering students can apply AI technology to an industrial setting. With AI Cup, we can spark creativity among students and also enable the effective implementation of AI in the industrial sector.”
Since its creation in 2018, TE AI Cup has developed into a global platform connecting engineering talent from Asia, North America and Europe to help solve real challenges. As part of the 2019-2020 competition, a team from South China University of Technology developed an unsupervised learning automatic detection system for the TE factory in Shunde, China. The system used AI machine vision technology to detect the appearance of plastic parts, ensuring there are no appearance defects for those products in the production line. This AI solution has been deployed at the Shunde site and was further upgraded to be able to complete the inspection of finished products on the production line. Compared to traditional inspection methods, the unsupervised learning automatic detection system can be deployed faster and is more efficient and accurate. As an example, this system can inspect a plastic component in only 0.33 seconds, with an accuracy rate of more than 99%.
Soochow University Professor Yihuai Wang, who is an advising professor for this year’s competition and has witnessed the platform’s development over the past three years, believes the program is much more than a competition.
"TE AI Cup gives university students an opportunity to participate in intelligent manufacturing upgrades and transformation and empowers them to be creative. The challenges faced by our industry are ever-changing, requiring us to seek more innovative solutions to meet the dynamic needs.” Professor Wang said.
This year’s winners were:
First Place
- Team Urameshi, Universidad de Sonora, Mexico
Second Place
- Team Venados, Instituto Technologic de Hermosillo, Mexico
- Team Xiamen University, Xiamen University, China
Third Place
- Team SWUFER, Southwest University of Finance & Economics, China
- Team AI SONORA, ITESM Campus Sonora Norte, Mexico
- Team The Firefly, Soochow University, China
