Darsh Gupta builds aircraft, radios, and the math between them.

Grade 12 student at NPS HSR, Bengaluru, working across UAV hardware, FPGA software-defined radio, electromagnetic simulation, and ML-assisted aerospace design.

View the build log Start a conversation

PCMCCBSE academic track

2021+self-directed coding

20+hands-on STEM courses

Q2 2026hardware test window

Night-lab snapshot Hardware, RF instrumentation, and design notes sharing the same bench.

Drone-v2 Airframe + avionics integration

Code-SDR FPGA and RF signal pipeline

darsh.workspace live public snapshot

DRONE_V2

Autonomous aircraft platform

Custom PCB, RF front-end, firmware, assembly underway.

CODE_SDR

Ultra-wideband SDR

FPGA DSP pipeline, 105 MSPS target, board routing and assembly.

RCS_ML

UAV stealth surrogate research

Regression models trained on EM simulation data for fast screening.

SSERD

Genex Space rover work

Internship thread moving from satellite documentation to rover systems.

01Current Build Pipeline

A public, portfolio-safe synthesis of what showed up across the site and Drive planning docs.

hardware

Drone-v2

Multi-layer autonomous aircraft board with sensor loops, telemetry, RF layout, and final assembly/test work.

rf + fpga

Code-SDR

High-speed software-defined radio: Verilog DSP, ADC interface, Ethernet streaming, and GNU Radio integration.

research

RCS surrogate

Paper on ML regression for UAV radar cross-section estimation, framed as a fast filter before full-wave validation.

systems

Rover work

SSERD/Genex Space internship direction, extending aerospace documentation work into rover systems practice.

Builder first, theory close behind.

I like projects where physics and code have to negotiate with real hardware: PCBs that need to be manufacturable, firmware that has to survive sensors being noisy, and simulations that should help decisions rather than just make pretty plots.

My Drive notes point to a deliberately broad technical arc: aerospace and mechanical engineering, RF systems, embedded control, AP Physics 2 / Calc BC / Chemistry preparation, SAT work, and college research across the US, Canada, India, and Australia.

The through-line is simple: learn the underlying math, build a working artifact, document the failure modes, then iterate.

Technical range

Python C / C++ Verilog Basic assembly OpenFOAM basics KiCad / PCB RF matching DSP CLI tools OOP / DSA

Public achievements

Skill Titans top-50 finalist Minvayu FabLab internship Astro Space Camp 2023/2024 HMUN + Techfest MUN cAIpher AI hackathon design

I used Drive content selectively: public-facing project, education, and activity signal made it in; private schedules, detailed scores, phone-style planning notes, and raw college planning rows stayed out of the page.

02Research Threads

The page now surfaces research as a serious part of the portfolio, not just a side note.

completed draft

Aerospace CFD and machine-learning research visualization

ML-Based Rapid RCS Estimation for UAV Stealth

Machine-learning surrogate models for estimating radar cross-section from EM simulation data, including dense nets, volumetric CNNs, transformers, PINN-style constraints, uncertainty, and full-wave validation strategy.

active / archived lessons

Engineering workspace with aircraft blueprint and RF instrumentation

AI-Based Fuselage Optimization

Generative aircraft design and CFD exploration. The plan notes also capture the honest engineering reality: some debugging paths get closed when the system stops being tractable.

next rough draft

Prototype aerospace hardware on a technical workbench

Hybrid-Electric Propulsion Sizing

Multi-objective evolutionary algorithms for balancing mass, efficiency, and emissions in aircraft propulsion design.