NASA IMPACT · AI4Science · NASA MSFC at The University of Alabama in Huntsville

Sujit Roy

AI foundation models for science.

I'm a Lead AI Researcher on IMPACT's AI4Science team at NASA's Marshall Space Flight Center (MSFC), at The University of Alabama in Huntsville, where I lead the development of foundation models for NASA's Science Mission Directorate — currently focused on heliophysics and planetary science. With NASA and IBM Research, my teams built Surya, trained on the full Solar Dynamics Observatory archive, and Prithvi — open Earth-observation and weather–climate models now used across the community — with lunar and Martian-atmosphere models underway.

I came to space science through the brain: a PhD in computer science (UKIERI fellowship) at Ulster University with IIT Kanpur, advancing MEG- and EEG-based decoding of motor imagery for practical brain–computer interfaces, then explainable-AI research at the University of Manchester. Across ten years of ML R&D — and as co-founder of BrainAlive Research — the throughline is the same: models that learn the dynamics of physical systems, from cortical signals to solar storms. Current interests: generative modeling, computer vision, explainable AI, and reinforcement learning.

NASA Early Career Achievement Medal
AGU Open Science Award
650k GPU-hour NAIRR allocation

Enter the system The autonomous scientist → Publications

Scroll on into the Sun · click a world to glide straight to it

05 · Phase space

Quantum kernels for chaos

Structured Quantum Kernels for Chaotic Forecasting · arXiv 2609.13360

AeRot: amplitude encoding of normalised delay windows plus one layer of single-qubit rotations that assigns contiguous blocks of time steps to each qubit. Kernel ridge regression on Lorenz-63, bandwidths cross-validated over 100 seeds, beats tuned RBF and Matérn-5/2 kernels at horizons of 0.15 time units and beyond. The gain is a finite-sample inductive-bias effect, localised to the unstable saddle-approach windows — not a computational speedup.

Chaos has a shape: two moments that look identical can be headed to opposite halves of it. A small quantum circuit, used as a similarity measure, keeps track of which way a trajectory is about to turn where ordinary methods blur the two together.

83%of the hardest windows won
5–8qubits
100seeds

Paper ↗

Quantum kernels · 1 / 4

Look-alike windows, opposite futures.

Near the saddle, two delay windows can be Euclidean-close yet commit to opposite lobes. Euclidean kernels average the branches — the grey path no trajectory ever takes. AeRot keeps the sign.

The gold and teal paths start almost on top of each other and end on opposite wings. Averaging them gives the grey ghost in the middle — a forecast of somewhere the system never goes.

Quantum kernels · 2 / 4

The win lives in the hard regime.

The advantage concentrates where the attractor is locally unstable: AeRot wins 83% of windows in the most unstable decile. Everything runs at 5–8 qubits — architectural, not a speedup.

Where the system is calmest, every method does fine. The quantum kernel earns its keep in the glowing band — the moments right before a turn.

Quantum kernels · 3 / 4

Next: a learned correction.

Forecast · observe · correct — no covariance, no noise model.

The forecast drifts, a few real measurements arrive, and a learned nudge pulls it back toward the truth. Then it drifts again, a little less each time.

Quantum kernels · 4 / 4

Structured Quantum Kernels for Chaotic Forecasting

Wang, Roy, Akbari, Maskey, Ramachandran · arXiv 2609.13360 · 2026

Read the paper ↗

06 · Deep space

Beyond the planets

Origins — a PhD in brain–computer interfaces at Ulster University with IIT Kanpur: decoding motor imagery from MEG and EEG in real time.

Where it all began — teaching computers to read imagined movement from brain signals, a path toward communication for people who can't move or speak.

MEG imagery dataset — open motor- and cognitive-imagery BCI dataset (Nature Scientific Data, 2021)
ALS detection — 3D deep CNNs on resting-state MEG; Franklin V. Taylor Memorial Award + IEEE Brain Best Paper (IEEE SMC, 2023)
Cortical decoding — mapping engagement during motor imagery, mental arithmetic & silent word generation (Human Brain Mapping, 2023)
Learning analytics — CLEFT engagement model (IEEE Access) and the AMBS AI toolkit at Manchester

Google Scholar ↗

Sujit Roy

Quantum kernels for chaos

Look-alike windows, opposite futures.

The win lives in the hard regime.

Next: a learned correction.

Structured Quantum Kernels for Chaotic Forecasting

Beyond the planets

An autonomous scientist

Selected publications

Invited talks & keynotes

Awards & funding

Gallery

Built together

Impact & coverage

Featured

Global press

Open models