Navya Autonomous Shuttle
Professional work on autonomous shuttle systems at Navya, focusing on motion planning, trajectory optimization, and real-time control for safe urban navigation.
Project Overview
Contributed to the development of planning and control systems for Navya’s autonomous shuttles, which operate in mixed traffic environments including pedestrians, cyclists, and other vehicles. The work focused on creating robust, safe, and comfortable navigation capabilities for public transportation scenarios.
Key Contributions
Obstacle-Avoidance Trajectory Planning
- Designed and implemented a real-time trajectory planner using Bézier curves for smooth path generation
- Applied numerical optimization techniques to generate collision-free paths around dynamic obstacles
- Integrated the planner into RTMaps middleware for real-time deployment on the vehicle
- Achieved sub-100ms planning cycles while maintaining passenger comfort constraints
Planning & Tracking Research
- Conducted comprehensive benchmarking of motion planning algorithms:
- Sampling-based methods: RRT, RRT* variants, Particle filter
- Graph search: Hybrid A* for structured environments
- Lattice planners: State-lattice for kinematic feasibility
- Model Predictive Control: Trajectory tracking and optimization
- Evaluated trade-offs between computational efficiency, path quality, and real-time performance
Model Predictive Control Enhancement
- Extended the MPC-based velocity control module for improved speed management
- Implemented predictive horizon optimization for smoother acceleration/deceleration profiles
- Integrated comfort constraints (jerk limits, lateral acceleration bounds)
- Improved energy efficiency through optimized velocity planning
Technologies Used
- C++ - Real-time control implementation
- RTMaps - Robotics middleware for sensor fusion and control
- MPC - Model Predictive Control for velocity profile generation
- Numerical optimization - Path planning and trajectory generation
- Python - Analysis, simulation, and benchmarking
Impact
The developed systems contributed to improving the safety, comfort, and reliability of Navya’s autonomous shuttles operating in real-world urban environments, serving thousands of passengers in deployment sites across multiple cities.
Skills Demonstrated
- Real-time motion planning and control
- Autonomous vehicle software architecture
- Safety-critical systems development
- Algorithm benchmarking and optimization
- Integration with production robotics middleware