Vijay Yadav

Isaac Sim Mobile Robot Control

A production-grade mobile robotics project demonstrating sim-to-real architecture with NVIDIA Isaac Sim 5.0 and ROS 2 Jazzy. The robot autonomously navigates a sequence of waypoints using closed-loop control with real-time odometry feedback. The control stack is fully decoupled from the simulator — the same code would drive a real Jetbot without modification.

Architecture

┌─────────────────────┐ ┌─────────────────────┐ │ ROS 2 Controller │ /cmd_vel │ Isaac Sim │ │ Node │ ──────► │ Standalone Script │ │ (Python 3.12) │ │ (Python 3.11) │ │ │ /odom │ │ │ │ ◄────── │ │ └─────────────────────┘ └──────────┬──────────┘ │ FastDDS bridge │ ▼ ┌──────────┐ │ Jetbot │ │ in Sim │ └──────────┘

Both sides follow the "functional core, imperative shell" pattern — pure math and state-machine logic are isolated from ROS 2 and Isaac Sim integration, making the core algorithms unit-testable without infrastructure.

Tech Stack

NVIDIA Isaac Sim 5.0 ROS 2 Jazzy Python 3.11 / 3.12 FastDDS colcon Ubuntu 24.04 LTS RTX 5090

🎥 Demos

Phase 2 — closed-loop waypoint navigation (left) · Phase 1 — basic motion control (right)

📂 Repository & Docs

🔗 GitHub Repo 📄 Architecture Doc 📘 README

📍 Roadmap

✅ Phase 1: Motion control with articulation-based wheel velocities
✅ Phase 2: Closed-loop waypoint navigation with /odom feedback and proportional control
🚧 Phase 3: Modular refactor complete; LiDAR + reactive obstacle avoidance in progress
📋 Phase 4: Nav2 stack, SLAM Toolbox, behavior trees
📋 Phase 5: Unit tests, Docker, GitHub Actions CI/CD

🤖 Other Work

sEMG-Based Gesture Control for Shadow Robot Hand

University of Hertfordshire · Jun 2024 – Sep 2024

Developed a real-time ROS-based control pipeline for a 25-DOF Shadow Robot Hand, integrating sEMG-driven ML inference for gesture-based manipulation.

Low-latency sEMG signal acquisition using the Myo armband
Designed and trained ML models (SVM, CNN, LSTM) for gesture recognition
Integrated ML inference directly into the control loop
Validated in Gazebo simulation and on physical hardware

ROS Shadow Robot Hand (25-DOF) sEMG PyTorch Gazebo

🌱 Vision

My long-term goal is to develop affordable, real-world robotics systems that can learn from their environment through observation and interaction.

I'm particularly interested in learning-driven robots that improve their behaviour over time using perception and feedback — with a focus on practical, everyday applications.

The aim is to reduce complexity and cost while enabling intelligent robotic systems that can operate effectively in real-world environments.

🚀 Featured Project