Thirty-two years of quiet
Over 32 years in the Indian Navy, my specialist core has been submarine acoustic stealth and underwater acoustics. For six years as Head of the Acoustic Stealth Group, Directorate of Submarine Design, I led platform- and equipment-level stealth design across India’s nuclear-submarine programme — developing India’s first indigenous submarine acoustic model, indigenising underwater radiated noise (UWRN) estimation, and institutionalising standardised shock-mounting design methods that reduced dependence on proprietary foreign software.
That work continues through Oravont Systems LLP, my defence-R&D and product practice. I also supervise doctoral and post-graduate research at the defence-technology interface, building on faculty tenures at DIAT and MILIT, Pune. I hold a PhD in Mechanical Engineering (Applied Machine Learning, DIAT 2018), a granted Indian patent in AI-based fault detection, and a sustained publication record spanning machinery diagnostics, AI in warfare, and quantum technologies.
Three tracks of work
Underwater acoustic stealth: technology & products
Submarine acoustic stealth engineering — underwater radiated noise control, signature management and shock mounting — supported by simulation tools: SSR‑Sim (sonar signal reflection), ABNContour (airborne-noise mapping) and SMS‑Suite (shock-mount simulation), for submarines and surface ships.
Deep learning for vessel detection & underwater domain awareness
Deep learning directly on raw passive-sonar waveforms: acoustic event detection, multi-class classification, and vessel fingerprinting / re-identification with propagation-robust embeddings — engagements spanning the NCIIPC AI Grand Challenge, iDEX DISC5 (Indian Navy), and a research collaboration in formation with IIT Delhi. The core architecture is SKANN, a raw-waveform Selective-Kernel Acoustic Neural Network (preprint, arXiv:2609.07399).
Quantum & emerging defence technologies
Authorship and research liaison in quantum sensing and free-space quantum communication; editor of The Beginner’s Guide to Quantum Computing (2024), an open edition of Hughes, Isaacson, Perry, Sun and Turner (Springer 2021, CC BY 4.0) at unlockqubits.com. Also generative AI (LLM + RAG) for engineering knowledge work.
From the bench
SKANN: open-set vessel re-identification from underwater ship-radiated noise
A raw-waveform Selective-Kernel Acoustic Neural Network that turns a hydrophone recording into a 512-dimensional hull fingerprint, matches it against an enrolled gallery, and rejects vessels it has not seen — with a cross-passage evaluation protocol that keeps the measurement free of leakage.
DEMON Analysis: Reading a Vessel’s Propulsion from Its Own Noise
The physics, the Hilbert envelope maths, reading the harmonic comb — and a case where this classical method caught a hidden defect in a synthetic AI training dataset.
Physics-Grounded Synthetic Underwater Acoustic Dataset
Openly released synthetic passive-sonar dataset for self-supervised pretraining — physics-based vessel signatures with validated modulation behaviour.
GitHub: Underwater-Acoustic-Synthetic-Dataset → · Read the article →