Before he came to Binghamton University to pursue his PhD, Alem Fitwi worked for years as the director of information and communications technology at Mekelle University in his native Ethiopia, overseeing the technology needs of more than 35,000 students, faculty and staff. He also worked on multimillion-dollar windfarm projects for multinational corporations, building sustainable power near Addis Ababa, the nation’s capital.
In 2017, his desire to earn his doctorate in computer engineering brought him to the U.S. and the Thomas J. Watson College of Engineering and Applied Science. He now calls Binghamton “a home away from home to me,” a place where he found not only an education but also personal support during one of the worst times in his life.
Growing up in Ethiopia, Fitwi found himself at the top of the class from “the moment I started school.” He earned a scholarship to a boarding school, which is where he discovered his love for computers.
“At the boarding school, there were games as well as some programming or basic applications like Microsoft Word, Excel and the like,” he said. “Those things helped me figure out my future. I was asked about what I would be in the future. I said I’d like to study electrical and computer engineering, particular computer engineering. I think everything started there.”
After high school, the Ethiopian Ministry of Education assigned Fitwi to Bahir Dar University for his BS degree, and after a couple of years in the workplace, he attended Addis Ababa University (his country’s oldest university) for his MS. His jobs at Mekelle University and windfarm projects followed.
So how did he end up on the other side of the planet, far from the life he knew? While Fitwi worked at Mekelle, the university’s president went on a trip to the U.S. and toured several educational institutions, including Binghamton. Fitwi heard great things about Watson College and its Department of Electrical and Computer Engineering, and he began to explore the ECE PhD program.
Although he had traveled to China and around Europe as part of the ICT, windfarm and other projects, Fitwi had never been to the U.S., but he felt like he had absorbed a lot of the culture through films.
“I didn’t have much interaction with my peers when I was a kid,” he said. “My parents gave me full freedom to do whatever I wanted, but at the same time I was very responsible and very ethical. The only thing I used to like doing apart from my education was going to the movies, so before I came here, I knew a lot about America already.”
His PhD research — with Associate Professor Yu Chen as his advisor — combined hot-button cybersecurity topics such as privacy, surveillance, deep learning, blockchain networks, cloud computing and edge computing. His dissertation, “Privacy-Preserving Surveillance as an Edge Service,” tries to strike a balance between public safety and the privacy of individuals.
“We have more than a billion CCTV cameras running around the globe today, and they record a lot of information about individuals without their knowledge and without any regard to their privacy,” he said. “If you are passing by any area under surveillance, they’re going to take your picture, and there is no such privacy protection.”
Fitwi’s idea, which he calls selective surveillance, is an artificial intelligence system that would recognize threatening actions or aggressive gestures and only then would start to record for the authorities. For instance, a security system would not collect images until someone pulls out a gun, and then a police operations center would be alerted to the problem.
One problem he needed to solve: To keep the device secure, he needed a program that would run on a device with limited capacity, not accessing other resources on a network until it was sure what it was seeing needed to be flagged.
“You have to design your algorithm to be lighter to fit into these resource-constrained edge devices,” he said. “You can come up with good algorithms that work fine on your computer, but when you reduce the size, the accuracy also goes down, so balancing things out was very challenging. It was a stressful thing at times, but we managed to find some workarounds. Everything is about compromise — you have to find the optimal points.”
During his time at Binghamton, Fitwi also worked as a graduate assistant for the Watson College Dean’s Office, writing programs to analyze student and faculty datasets to make them more easily understood, and he also prepared surveys when needed for feedback.
Fitwi said he’s “always been a happy man,” but that optimistic outlook was tested in early November 2020 when the Ethiopian National Defense Forces clashed with the Tigray People’s Liberation Front, sparking a civil war in his country. The fighting has left thousands dead and displaced more than 2 million people from their homes.
“Back home, we are very much connected — the family connection is very, very, very strong,” he said. “But I was not able to communicate with my family for about 70 days after the war broke out. I’d like to say thank you to the ISSS [International Student and Scholar Services] office, the Watson Dean’s Office and others who tried to comfort me during the most difficult part of my life.”
Fitwi earned his PhD in December but is returning to campus this week to take part in Commencement. Since the start of 2022, he has worked as a software and algorithm engineer for Meta, the parent company of Facebook and Instagram. Although he can’t say exactly what he’s working on, it is related to image processing.
Having worked with large corporations in the past, Fitwi is not intimidated by them but also feels grateful to be part of one that has a global reach. He also has much gratitude for Chen, his PhD committee, the Watson College Dean’s Office and others who helped him at Binghamton.
“I would like to say thank you to the entire community of Binghamton University,” he said. “I had the most interesting experience of my life. I don’t know if I can say thank you enough, because it went beyond my expectations. I wish to come back to that place one day.”
Posted in: Science & Technology, Watson
