Issue link: https://iconnect007.uberflip.com/i/1520492
166 I-CONNECT007 I REAL TIME WITH... IPC APEX EXPO 2024 SHOW & TELL MAGAZINE Soroosh's primary role is to manage the col- lection, organization, and cleaning of data from these experiments, then develop pre- dictive models using machine learning tech- niques to forecast failures of solder joints and analyze the impacts of various factors, such as voids and other parameters, on the life and failure modes of solder joints. This project, he said, is currently supported by the U.S. Department of Defense. Soroosh believes the results of his research will save time and money for PCB companies. "I am developing machine learning mod- els that can predict the failure of solder joints with nearly 94% accuracy," Soroosh said, adding that he's actively working to enhance their accuracy and applicability to other tests, various solder alloys, component types, and additional factors. "The models I'm creating are capable of predicting the lifespan of solder joints for components and alloys that have not been previously tested, eliminating the need for extensive testing periods which can span months or years." Finally, he said his prediction process is extremely fast, "taking only milliseconds to estimate the lifespan of electronic compo- nents once the relevant data is inputted." He believes it will pave the way for more innovative and reliable electronics design and manufacturing. As for Palash, the importance of his research lies in addressing the knowledge gap between drop shock tests predominantly performed at room tem- perature and tests at elevated temperatures. "Performing tests at room temperature fails to accurately simulate real-world condi- tions where electronic circuits contend with operational or environmental thermal strains during normal operations," he said. "We want to ensure the reliability of solder joints in practical applications." Attending IPC APEX EXPO 2024 and having the opportunity to present their poster has been a highlight of their academic experi- ence. "It was a great experience interact- ing with the attendees at the poster session and explaining to them about my research," Palash said. While at the show, they attended keynote speakers, Technical Conference sessions, and Professional Development Courses. Palash said he particularly enjoyed "Intro- duction to Machine Data Analytics in the EMS Industry" by Tim Burke. "This course familiarized me with various aspects of machine data analysis, such as the available tools and software, including Grafana and Kibana," Palash said. Another highlight was attending technical sessions on AI and ML, particularly as they relate to X-ray inspection and void inspection in solder joints. He spent quite a bit of time at the OMRON booth, watching an entire demo of their machines, and learning about the latest SPI, AOI, and AXI systems. Palash's Interview