LiDAR-based AR Fluid Simulation

The Anisotropic App

Making invisible airflow legible in physical space.

Problem

Fluid behavior is difficult to build intuition for because velocity fields, pressure gradients, and transport mechanisms are invisible. Classroom demonstrations and textbook diagrams abstract away the spatial structure that makes flow problems genuinely anisotropic.

Technical Approach

The lab is developing a LiDAR-based AR tool that captures room-scale geometry via ARKit mesh reconstruction, then overlays a lightweight real-time CFD approximation mapped onto the captured surface. The stack is Swift on iOS, using LiDAR depth data to define boundary conditions and a simplified solver loop tuned for interactive frame rates rather than full Navier–Stokes fidelity.

Concept / prototypeSwiftARKitLiDARReal-time CFD approximation
AR preview domain — LiDAR mesh — flow overlay
Intended Use Case

Education and intuition-building for fluid behaviour.

Built for students and early researchers who need mechanistic intuition for fluid behavior — vent placement, boundary-layer effects, and directional transport — without standing up a full simulation pipeline.

Additional Projects

Engineering and research tooling.

Healthcare Platform

Hamara Prayas iOS App

Real-time blood donation requests and verified bank locators.

Hamara Prayas addresses chronic blood shortages for Thalassemia and Leukemia patients in Bihar, born from direct experience with gaps in availability during critical treatment periods.

The lab developed an iOS application in Swift featuring real-time blood requests, verified blood bank locators, and live community alerts. The platform has supported 24 blood donation camps collecting 6,956 units and reaching 300+ patients.

IIT Roorkee Internship Tooling

Automated CFD Post-Processing Pipeline

Python automation for combustion and exhaust simulation outputs.

Large CFD campaigns in hydrogen combustion and exhaust thermodynamics generate hundreds of field snapshots that require consistent extraction, normalisation, and comparison across inlet configurations.

Built during the IIT Roorkee research internship, this pipeline automates post-processing of Ansys Fluent and OpenFOAM outputs in ParaView — including H₂O mass-fraction stratification metrics and Tetens-formula saturation ratio fields across 10 inlet configurations.