By Sherry DiBari
At Hitachi America in Santa Clara, California, two Monarchs – one Sri Lankan, one Bengali – often get together for coffee. To get there, the two research scientists separately crisscrossed the world from South Asia to Norfolk, Virginia, then to California.
“I see Lasitha (Vidyaratne) almost every day,” said Mahbubul Alam, ’18 Ph.D. “We worked together in the ODU Vision Lab for more than six years. Apart from working on multiple research projects and publishing more than 20 research articles together, we became very good friends.”
“It definitely creates a sense of ODU community in the workplace,” Vidyaratne added.
Mahbubul (pronounced Ma-bu-bool) Alam and Lasitha Vidyaratne ’20 Ph.D., both graduates of the Department of Electrical and Computer Engineering (ECE), came to ODU by chance.
Alam, a senior research scientist, grew up in Dhaka, Bangladesh, with a picture-perfect childhood.
“The place where I used to live in my childhood was inside a university campus which was full of trees, open fields and lots of beautiful lakes,” Alam said. “It was completely a safe environment, and hence, I got the opportunity to explore outside a lot on my own without my parents being worried about me. I used to spend hours climbing into different trees, catching fish with my friends and playing in the open fields. The natural beauty of the place still makes me homesick from time to time.”
It was an academic atmosphere as well. His father is a physics professor and university president. Alam’s uncle, also a physics professor, is one of the lead dusty plasma scientists in the world.
Alam came to Old Dominion University sight unseen. One of his childhood friends, Sabah Zeehan Mirza, did her undergraduate studies at ODU.
“She spoke highly of ODU and its beautiful culture,” he said. “I applied to five or six schools and got admission to most of them but decided to come to ODU because of all the nice things I heard from her.”
Vidyaratne grew up in Colombo, Sri Lanka, an area acclaimed for its “Golden Mile” of tourist beaches. However, Vidyaratne’s childhood was shaped by decades of civil war.
He had several family members who were engineers or in other STEM fields. “I was probably getting a lot of STEM influence all the time,” he said.
He worked as an instructor for a while after earning his MEng in electrical and computer engineering in Malaysia. But Vidyaratne’s goal had always been to earn his doctorate and work in research and development.
“Ever since I was an undergrad, I was very interested in research in the fields of artificial intelligence and machine learning for biomedical and computer vision applications,” he said.
Searching for a place to earn his Ph.D., he came across Khan Iftekharuddin's research on the internet.
“I personally wrote to Dr. Iftekharuddin, and he encouraged me to apply for a Ph.D. at his lab based on my educational qualifications, work and research experience, and research interests,” Vidyaratne explained.
Iftekharuddin mentored Vidyaratne and Alam in ODU’s Vision Lab. They thrived in the research environment.
In 2017, the duo, along with another doctoral student, Zeina Shboul, placed first out of 17 teams for their work on evaluating brain tumors using magnetic resonance imaging (MRI) scans.
“Both Mahbub and Lasitha worked on an NSF-funded project for study of novel brain-inspired AI/machine learning methods for invariant pattern recognition in the Vision Lab,” Iftekharuddin said. “The Vision Lab approaches research from dual focus on fundamentals and applications for any specific topic. Following this approach, they both made significant contributions in addressing a few of the long-standing challenges in high-performance computing with limited labeled data and time series data. As for applications, they focused on different data science applications: one on security and the other on biomedical. They really worked well as a team and helped train incoming graduate and undergraduate students in the lab.”
At ODU, Alam’s research was centered on artificial intelligence.
“My dissertation focused on proposing more efficient and robust AI techniques by addressing some of the limitations of the existing AI models, such as use of excessive computer resources,” he explained.
He also mentored undergraduate students with research projects, including designing hardware accelerated AI models, designing and implementing AI models in the NAO humanoid robotic platform, and applying novel AI techniques in the wireless network domain.
Vidyaratne’s dissertation at ODU was on inventing a more efficient artificial neural network architecture specially for processing multichannel frequency data, such as information from medical sensors and machine health signals.
As a postdoc at Jefferson Lab, Vidyaratne worked on a Department of Energy-funded project to develop machine learning models to automatically identify RF faults (fault type and the component that failed) in superconducting radio frequency cavities within the continuous electron beam accelerator facility (CEBAF).
When Alam’s lab at Hitachi started hiring new researchers, he referred Vidyaratne.
“He got in after successfully passing multiple rounds of tough interviews,” Alam said. “Having Lasitha around and working together again created a great sense of ODU community as we share a lot of common interests. We hope to do a lot of great research works together again.”
At Hitachi, Vidyaratne works with machine learning.
“I build machine learning-based solutions for various industry requirements such as maintenance and repair analytics, operation optimization and quality enhancement in order to improve the efficiency and production capability of major industries worldwide, such as transportation, mining, manufacturing and energy,” he explained.
Alam works in the Industrial AI Lab.
“We design innovative AI-based solutions for many industry applications,” he explained. “One of the notable recent works that I have done is designing a one-of-a kind AI solution for one of the largest fleet-management companies, Penske, for estimating breakdown time of the thousands of Penske trucks/fleets. Penske is using this solution in the field, and it already saved millions of dollars within a very short period of time and hopes to save a lot more in the future. I was the lead researcher and the technical lead of this extremely sensitive and important project. Such a sophisticated solution has never been used by any companies in the field at large.”
Alam and Vidyaratne may not be the only ODU Vision Lab alumni to end up at Hitachi. Alam referred ECE graduate student Russell Kazi for an internship during the winter of 2020.
“He did a great job,” Alam said. “I always try my best to refer as many ODU students as possible and try to advertise ODU as one of the best schools in the United States. ODU is and always will be one of the greatest places I have ever been.”