The Red Hat Collaboratory at Boston University is supporting select undergraduate student research projects during Summer 2022, in keeping with its mission of advancing education in open source technologies. So far, six projects have been chosen to receive funding and supervision from BU computer engineering professors active in their own Collaboratory projects, with more expected.
The award recipients were selected from a competitive pool of applications in areas including cloud computing, systems engineering infrastructure, and security:
- Nengneng Yu, “Fine-grained, automated security detection via semi-supervised learning,” in collaboration with PhD student Yajie Zhou, supervised by Professor Alan Liu. The project aims to detect and recover attack stories behind security incidents using natural language processing techniques.
- Julia Hua, “Network-accelerated in-memory key-value store live migration,” in collaboration with PhD student Zeying Zhu, supervised by Professor Alan Liu. The project aims to develop an efficient open source software migration system to meet demands for high throughput and low latency in the cloud.
- Quan Pham, “Real-time quality assurance,” supervised by Professor Gianluca Stringhini. The project aims to develop plugins for the Jupyter Notebook environment that will analyze code and provide real-time feedback on software in development.
- Ethan Klein, “Red Hat unikernel Secrecy project,” supervised by Professor Orran Krieger. The project aims to develop a unikernel implementation of Secrecy, a platform enabling parties to perform shared computations on private data without sharing their actual data.
- Xiteng Yao, “Practical programming of FPGAs with open source tools: test-code generation and guided search through supervised learning,” supervised by Professor Martin Herbordt. This project aims to improve an open source programming tool flow known as LLVM using reinforcement learning.
- Shun Zhang, “D-COLLECTIVE: democratized data collection and collaborative training for extreme-scale autonomous systems,” supervised by Professor Eshed Ohn-Bar. This project aims to develop technologies to allow everyday people to participate in data collection and model training for autonomous systems, specifically self-driving cars.
Student award recipient Xiteng Yao heard about the Red Hat Collaboratory from Professor Herbordt, who has worked on multiple projects as part of the Collaboratory. “I was excited to see so many interesting projects in the Red Hat Collaboratory,” Yao said. “It’s interesting to work in a team with academic and industry members; it is an excellent opportunity to learn from different types of people.” Open source research was another appealing aspect of working with the Collaboratory: “What’s exciting about open source research is that your work can be reused by numerous developers around the world. I’d like to develop a robust solution to a real- world problem that could help others.”
BU professor Ari Trachtenberg was instrumental in developing the program and says that undergraduate researchers are essential to the long- term viability of the Collaboratory’s mission. “Not only can these students make valuable contributions in their own right,” Trachtenberg said, “but they will also become the graduate students, engineers, and professors that will drive the computer systems community going forward. It’s important that these students experience practical research mentorship to preserve the cultural aspects of modern systems research, such as properly establishing, controlling, and analyzing large-scale tests; determining convincing evidence of improvement; and understanding the subtleties involved with reading or writing a technical systems paper.”
“I think the most important feature of this program is that it has enabled computer systems research that may otherwise not have occurred.”
—Prof. Ari Trachtenberg
Systems research typically requires significant manual and infrastructural investments from a wide variety of groups over a sustained period, making it out of reach for most undergraduate researchers.
The Red Hat Collaboratory and its various partners, such as the MOC Alliance, give researchers a ready- made experimental environment and a collection of interested colleagues. This setting provides the students and their mentors the opportunity and the means to try out new systems ideas without the heavy lift of a national, multi-partner grant. “I think the most important feature of this program is that it has enabled computer systems research that may otherwise not have occurred,” Trachtenberg said.
Undergraduate research projects are generally projects that can be completed within a shorter time frame, such as a four-month summer session, by a junior researcher with limited experience. Some of the projects are also extensions of projects that the Collaboratory is already funding, such as projects relating to unikernel Linux, practical programming of FPGAs, and security detection.