NSF Recognizes Six Boston University Researchers with CAREER Awards for Advances in Trustworthy AI, Robotics Safety, Data Systems, and More

Six Boston University researchers have been awarded NSF Faculty Early Career Development (CAREER) awards, recognizing their leadership as emerging scholars and their potential to shape future research and education across computing, engineering, and data science. The 2025 cohort’s projects span trustworthy healthcare AI, equitable public resource allocation, large-scale data systems, robotics safety, accessibility-focused AI, and foundational computational theory, all areas with significant implications for security, privacy, and societal resilience.

“These awards reflect the extraordinarily talented and creative faculty that Boston University is able to recruit,” said Kenneth Lutchen, BU vice president and associate provost for research. “Each of these projects reflects the impact-driven scholarship that defines our research community.”

Below are highlights of the six CAREER-funded projects:

1. Making Medical AI More Transparent and Trustworthy

Kayhan Batmanghelich (ENG assistant professor of electrical and computer engineering)

Many medical AI systems still function as opaque “black boxes,” complicating clinical adoption and eroding public trust. Batmanghelich will develop methods to convert these systems into human-interpretable models, translating predictions into logical rules or simplified programs. The project will also explore using AI to audit other AI systems, generating accessible explanations for diagnostic models in areas such as breast cancer and chronic lung disease.

Impact: Supports the development of equitable, explainable, and reliable AI for high-stakes medical applications.

2. Fairer and More Efficient Allocation of Social Services

Kira Goldner (Computing & Data Sciences assistant professor)

Goldner studies the economic and computational trade-offs behind the “ordeals”, long waits, paperwork, verification processes, that often govern access to public services. Her CAREER project seeks simple, explainable mechanisms that more equitably ration services without imposing unnecessary burdens on consumers.

Impact: Could reduce inequities in social service delivery and inform evidence-based public sector policy.

3. Scalable, Accessible Data Systems for Society

Vasiliki Kalavri (CAS assistant professor of computer science)

Kalavri develops systems that make it possible for nonexperts and small organizations to analyze large volumes of data from wearables, sensors, vehicles, and IoT devices. Her work focuses on building efficient, secure, and user-friendly analytics platforms that support applications in smart cities, digital health, disease surveillance, and environmental monitoring.

Impact: Empowers communities and researchers with tools for real-time, data-driven decision-making.

4. Approximating Solutions to NP-Hard Problems

Nathan Klein (CAS assistant professor of computer science)

Klein’s research tackles classic NP-hard optimization problems, like the traveling salesperson problem, by studying how close approximation algorithms can get to the optimal solution. These algorithms are essential for real-world challenges, from logistics to circuit design.

Impact: Advances theoretical computer science while supporting practical applications across infrastructure, mobility, and operations.

5. Designing Safer, More Efficient Robots

Sabrina Neuman (CAS assistant professor of computer science)

Neuman aims to automate the creation of customized computing architectures that help robots plan movement more safely and efficiently, using information about each robot’s physical structure. Her focus includes applications in assistive care, industrial automation, and healthcare environments.

Impact: Supports safer human–robot interaction and more energy-efficient robotic systems.

6. Accessible AI for Users with Low Vision

Eshed Ohn-Bar (ENG assistant professor of electrical and computer engineering)

Ohn-Bar will develop AI systems that better reflect the needs and preferences of people with vision impairments. His team will build a dataset of real user interactions and design models that learn directly from user feedback, addressing issues like directional misinterpretation and inaccurate descriptions.

Impact: Sets new standards for inclusive, user-centered AI, with potential benefits for navigation systems, smartphones, and smart city technologies.

These projects highlight BU’s growing strength at the intersection of AI, human-centered computing, data systems, and societal impact, while offering significant opportunities for student research and training through the NSF CAREER program.

Submitted by Jason Gigax on Fri, 11/14/2025 - 09:39