Student researcher · Aerospace · ML · Quantum

I study how things move, think, and cannot be known.

I'm Neil Akhawat — a student researcher working at the intersection of hypersonic flight, machine learning, and quantum mechanics. My current focus is real-time trajectory prediction for hypersonic glide vehicles: estimating where a Mach–15 vehicle will be when classical tracking breaks down.

View research → About me ✉ Subscribe

Focus

Hypersonic trajectory prediction

Fields

Aerospace · ML · Quantum

Writing

neilakhawat.com

Video

@neilakhawat

01 · Who I am

About me

A student researcher trying to understand the hardest things to predict — vehicles that evade prediction, systems that learn, and a universe that is uncertain at its core.

Neil Akhawat

Student Researcher

FOCUSHypersonic trajectory prediction

FIELDSAerospace · ML · Quantum

WRITINGneilakhawat.com

VIDEO@neilakhawat

I'm Neil Akhawat, a student researcher fascinated by the boundary between what we can compute and what we can never quite pin down. My work sits where aerospace, machine learning, and quantum mechanics meet — three fields bound together by a single question: how do you predict something that resists being predicted?

Right now that question takes a very concrete form. A hypersonic glide vehicle flies at Mach 5–25, maneuvers on its way down, and wraps itself in plasma that blinds radar at exactly the moment tracking matters most. I'm building a physics-constrained pipeline that reconstructs a vehicle's hidden aerodynamic state from noisy sensor data and forecasts where it will go — a calibrated confidence corridor rather than a single guess.

Around the core research, I write to make hard ideas legible — from missile flight regimes to Heisenberg's uncertainty principle — and I make video explainers that break these topics down for anyone curious about how the physical and computational world actually works.

I believe the most interesting problems live at the edge of certainty, and I'm building the tools — and the intuition — to work there.

Hypersonics Trajectory prediction State estimation Physics-informed ML Sequence models Quantum mechanics Scientific writing

My research → Get in touch

02 · What I work on

Research areas

My work spans three connected questions: how do we predict the path of a vehicle that actively evades prediction, how do learning systems model physical processes, and how do we reason under fundamental uncertainty. The common thread is prediction under hard constraints.

● Active · Primary

Aerospace · Hypersonics · Defense

Hypersonic glide vehicle trajectory prediction

A hypersonic glide vehicle travels at Mach 5–25, maneuvers actively, and ionizes the air around it — disappearing from radar exactly when tracking matters most. My research builds a physics-constrained prediction pipeline that reconstructs a vehicle's hidden aerodynamic state from noisy radar and satellite data, forecasts its future maneuver intent rather than its raw position, and propagates trajectories through a digital twin built on hypersonic equations of motion and a standard atmosphere model. The goal is a calibrated confidence corridor for where the vehicle will be — generated from realistic sensor input alone, surviving plasma blackout.

IMM-UKF state estimation Intent prediction Plasma blackout modeling Physics-informed ML Anderson hypersonic EOM Monte Carlo corridors

● Ongoing

Machine Learning · AI

Sequence models & learning under structure

How do recurrent and attention-based models learn temporal patterns — and how do we keep them physically honest? I work with dual-channel recurrent networks, sparse causal attention, and physics-residual training so that learned predictions never violate the laws governing the system they model.

Bi-GRU / LSTM Sparse attention Physics-residual loss

● Exploratory

Quantum Mechanics

Uncertainty & quantum machine learning

From the Heisenberg uncertainty principle to the question of how quantum systems can accelerate learning — I explore where quantum mechanics meets computation, and what it means to predict in a world that is fundamentally probabilistic at its base.

Uncertainty principle Quantum ML Probabilistic reasoning

● Foundational

Missile Systems · Flight Dynamics

Missile trajectories & flight regimes

The groundwork for the hypersonic research: understanding how ballistic, cruise, and glide missiles differ by their trajectories, the physics of each flight regime, and why each demands a different approach to tracking and interception.

Ballistic vs glide Flight dynamics Trajectory analysis

03 · Community

Aerospace × Artificial Intelligence Discussions

Ask a question, answer someone else's, and dig into the ideas behind flight, learning, and the physics underneath. Search first — your question may already have an answer.