IndiMorph
From Deserts to the Himalayas

Bharat’s Own Next Generation AI Powered Shape Shifting Mobility An indigenous multi modal platform forged for the land of extremes from the blazing sands of the Thar to the icy heights of the Himalayas. Powered by cutting edge AI and edge intelligence, it morphs in real time to serve the nation in transport, defense, and disaster relief.

IndiMorph Platform

System Overview

IndiMorph integrates state-of-the-art deep learning models, real-time control logic, digital twin visualization, and cloud-native DevOps practices for rapid innovation in smart mobility.

AI Models

Terrain classification, 3D CAD generation, and CFD optimization using EfficientNetV2, MobileNetV2, and reinforcement learning for adaptive morphing decisions.

Edge Controller

Lightweight inference pipeline on Jetson Nano/Raspberry Pi with real-time telemetry logging, PID control, and LSTM-based health monitoring.

Hardware Control

Arduino firmware for Nitinol wire and pneumatic system control with Python serial bridge for robust, thread-safe communication.

Digital Twin

Unity-based real-time simulation synchronized with physical and AI states via MQTT for live visualization and operator control.

Cloud & DevOps

Prometheus monitoring, Grafana dashboards, JWT authentication, and Terraform deployment automation for scalable cloud infrastructure.

Datasets & Simulations

Structured data logging, OpenFOAM CFD simulations, Unity batch runners, and cloud storage integration for research reproducibility.

AI Models

Advanced machine learning modules for terrain perception, 3D shape generation, and aerodynamic optimization enabling intelligent morphing decisions.

Terrain Classifier

EfficientNetV2 and MobileNetV2 models for real-time terrain classification using camera and IMU data fusion.

CAD Generator

VQGAN3D/UNet/Transformer3D architectures for 3D vehicle shape synthesis and design exploration.

CFD Optimizer

Reinforcement learning with OpenFOAM integration for aerodynamic optimization and morphing pattern discovery.

Hardware Control

Robust actuator management and sensor integration for reliable shape-shifting capabilities across diverse operational environments.

Arduino Firmware

Real-time servo control, serial command parsing, and safety validation for Nitinol wire and pneumatic system actuation.

Python Bridge

Thread-safe communication layer with configurable serial ports and extensible command API for hardware integration.

Morph Logic

Decision-making engine translating AI outputs into actuator commands with context-aware pattern mapping.

Health Monitoring

LSTM-based anomaly detection for predictive maintenance and fault detection in actuator and sensor systems.

Digital Twin

Real-time simulation and visualization platform providing live digital twin capabilities for operators and researchers.

MQTT Integration

Real-time data streaming from physical and AI systems to Unity visualization with low-latency, high-throughput communication.

Unity Visualization

Rich, interactive 3D visualization with real-time updates reflecting morphing, terrain changes, and sensor readings.

Twin Sync Interface

UDP socket bridge and MQTT relay for bidirectional communication between backend systems and Unity digital twin.

Scenario Simulation

Batch and scenario-based simulation for research and testing with comprehensive logging and analysis capabilities.

Example Workflows

End-to-end morphing cycles demonstrating the complete system integration from sensing to actuation and monitoring.

1

System Startup

Vehicle powers on, initializing all hardware, edge controllers, and backend services with health checks and calibration.

2

Terrain Classification

Edge controller runs TFLite terrain classifier on live camera and IMU data, outputting current terrain type for morphing decisions.

3

Morph Decision

Morph logic module receives terrain type and mission mode, computes optimal actuator pattern, and sends commands via Python serial bridge.

4

Actuation

Arduino firmware actuates Nitinol wires and pneumatic systems, morphing the vehicle shape according to computed patterns.

5

Digital Twin Update

Current state published via MQTT and visualized in Unity, providing live digital twin for operators and researchers.

6

Monitoring & Control

LSTM health model monitors for anomalies while operators can view and control the system via web dashboard.

Developer Team

Meet the brilliant minds behind IndiMorph a diverse team of researchers, engineers, and innovators driving the future of AI powered adaptive mobility.

OM SINGH

OM SINGH

AI Researcher and Developer

Specializes in deep learning and computer vision with advanced experience in autonomous systems.

GAURAV PANDEY

GAURAV PANDEY

AI Researcher and Developer

Specialist in AI driven perception and control systems with expertise in shape shifting mechanisms.

ANUSHKA DWIVEDI

ANUSHKA DWIVEDI

Full Stack Developer & Designer

Full stack developer specializing in cloud infrastructure and real time data processing systems.

JAHNAVI SHUKLA

JAHNAVI SHUKLA

UI/UX Designer

Specializing in intuitive interfaces, User centered design, and seamless experience for smart mobility platforms.

MANYA SRIVASTAVA

MANYA SRIVASTAVA

Full Stack Developer

Full stack developer specializing in cloud infrastructure and real time data processing systems.

YASH SRIVASTAVA

YASH SRIVASTAVA

AI & Embedded Systems

Specializes in real-time control systems and embedded AI, contributing to the integration of deep learning models.

Research Partners

Collaborating with leading institutions and organizations to advance AI-powered adaptive vehicle technology.