
Aman Patel
Robotics Expert
Kompetenzen

Meine Dienstleistungen


Portfolio
Arbeitserfahrung
Industrial Automation
ABB • Vollzeit
Sep 2024 - Present • 1 yr 9 mos
1. Robotics & Simulation (ROS2/MoveIt2): I design and program high-fidelity robotic simulations using ROS2 and Gazebo. My expertise includes motion planning with MoveIt2, URDF modeling, and developing obstacle avoidance logic for industrial-grade robotic arms. I apply advanced kinematic modeling to ensure precision in dynamic environments. 2. Data Analytics & Competitive Intelligence: Leveraging Python (Pandas/NumPy), I engineer custom analytical tools, including a proprietary Fiverr Competitor Analysis Dashboard. I help sellers and businesses transform raw market data into actionable growth strategies through statistical modeling and SEO trend analysis. 3. Custom Web Architecture: I develop performance-optimized, scalable websites from scratch. Unlike template-based solutions, I focus on clean, custom-coded architectures that ensure maximum speed, responsiveness, and SEO visibility. Technical Stack: Languages: Python, C++, JavaScript Robotics: ROS2, MoveIt2, Gazebo, URDF Data: Pandas, NumPy, Scikit-learn, Statistical Analysis Web: Full-stack development, SEO Strategy, Custom Frameworks
Robotics Simulation & AI Developer
Robotic Imaging • Freiberufler
Jan 2023 - Sep 2024 • 1 yr 8 mos
1. High-Fidelity Physics Simulations (NVIDIA Isaac Sim / MuJoCo): Developed GPU-accelerated digital twins for industrial robotic arms and quadrupeds. Processed CAD assemblies and converted URDF/OpenUSD files, adjusting PhysX collision meshes, inertia tensors, and joint limits for realistic contact dynamics. 2. Reinforcement Learning & Motion Planning: Designed custom Gym environments in Isaac Lab. Formulated reward functions, observation spaces, and action bounds to train robust PPO locomotion and manipulation policies using PyTorch. Automated training scripts on headless GPU cloud servers. 3. Sim-to-Real Deployment & Docker: Created ROS2 communications bridges to transfer trained AI policy weights from simulation to physical controllers. Containerized complete software workspaces inside Docker to eliminate local dependency issues and ensure cross-platform reproducibility. Technical Stack: - Simulation: NVIDIA Isaac Sim, Isaac Lab, MuJoCo, Webots - AI Frameworks: PyTorch, Stable-Baselines3, rsl_rl - Systems: ROS2, C++, Python, Docker, CUDA