Seven EE Faculty Receive Columbia Research Stabilization Funds

Seven Columbia Electrical Engineering faculty members have been awarded funding from Columbia University’s Research Stabilization Fund, a university-wide initiative designed to support ongoing research disrupted by funding gaps. Administered by the Office of the Executive Vice President for Research, the stabilization initiative aims to maintain continuity for faculty, students, staff, and labs by offering bridge support as researchers seek new funding sources or pivot their work in response to emerging needs.

The faculty members—Micah Goldblum, Homayoon Beigi, James Anderson, Savannah Eisner, Zoran Kostic, Ethan Katz-Bassett, and Christine Hendon—represent a diverse range of research fields spanning artificial intelligence, civil infrastructure, sustainable energy, healthcare, and machine learning.

Micah Goldblum
Goldblum’s lab will use the funding to train models that assess the likelihood of AI agents engaging in risky behavior. The project aims to develop tools that guide AI systems toward safer behavior through pre-emptive modeling—laying the groundwork for a new class of guardrail technologies for artificial intelligence.

Homayoon Beigi
Beigi is collaborating with professor Raimondo Betti from Civil Engineering on a project titled From Single-Component Laboratory Experiments to In-Service Assessment: a New Paradigm for Machine Learning in Structural Health Monitoring of Civil and Mechanical Systems. The funding will support a new postdoctoral researcher to expand their machine learning models for real-time infrastructure monitoring, using lab-generated data to simulate live bridge and building conditions.

James Anderson
Anderson’s project, Closed-loop Conformal Decision Theory, aims to improve how energy storage operators make decisions in volatile electricity markets. The research focuses on energy arbitrage—charging batteries when electricity prices are low and discharging when prices are high—to generate profit. While such decisions are typically based on price forecasts, those forecasts often contain uncertainty, which can lead to unprofitable charging schedules. To support smarter, more risk-aware decisions, the project will embed conformal decision theory—a statistical uncertainty quantification methodology—into a closed-loop real-time feedback system to improve profitability in highly volatile markets.

Savannah Eisner
Eisner’s project, Seed Funding for a Fusion Diagnostics Research Program using Radiation-Hardened Semiconductor Sensors, supports a strategic pivot into diagnostics for fusion energy systems. Her team will develop advanced Hall-effect sensors and x-ray detectors designed to withstand the extreme conditions of fusion reactors, such as high radiation, heat, and electromagnetic interference.

The funding will support a graduate research assistant and new equipment for testing in her lab and at the Columbia Plasma Physics Laboratory’s High-Beta Tokamak (HBT). “Without stabilization funding, we cannot generate the early validation data needed to activate external collaborations,” Eisner said.

Zoran Kostic
Kostic is advancing research in low-power AI by developing methods for knowledge distillation that allow high-capacity models to be compressed into forms suitable for healthcare and IoT devices. His team is building tools for audio-based differential diagnosis, with applications in patient monitoring and remote care.

Ethan Katz-Bassett
Katz-Bassett will conduct a feasibility study exploring whether Internet measurement tools can help verify the physical location of AI chips, which play a critical role in machine learning. His work responds to proposed federal legislation that seeks to ensure compliance with export controls by tracking the use of high-performance chips.

Christine Hendon
Hendon’s research bridges engineering, pathology, and gynecologic oncology to improve early detection of endometrial cancer. Her team is developing a hyperspectral imaging platform, enhanced by machine learning, to build a spectral tissue database. The aim is to enable earlier, noninvasive, and more accurate diagnostic tools.

“This work unites engineering, pathology, and gynecologic oncology in a unique collaboration,” Hendon said. “By building a spectral map of the uterus and how it changes with disease, we aim to open the door to noninvasive tools that can detect cancer earlier and more accurately.”

The Research Stabilization Fund is not intended to replace federal funding but instead acts as a bridge to help researchers complete projects, pursue alternative sources of support, or explore new directions. It represents an investment in our exceptional faculty and in our mission