Table of Contents

Main Sections

1. Structured Learning Path 2. Major Algorithms, Techniques, and Tools 3. Cutting-Edge Developments (2023-2025) 4. Project Ideas by Level Recommended Learning Strategy Essential Resources

Learning Phases

Phase 1: Mathematical Foundations Configuration Space & World Representation Phase 3: Classical Path Planning Algorithms Phase 4: Sampling-Based Planning Phase 5: Trajectory Optimization Phase 6: Navigation for Specific Robot Types Phase 7: Dynamic and Real-Time Planning Phase 8: Localization and SLAM Phase 9: Advanced Topics

Project Levels

Beginner Projects Intermediate Projects Advanced Projects Expert/Research Projects

Complete Roadmap for Path Planning and Navigation in Robotics

1. Structured Learning Path

Phase 1: Mathematical Foundations (4-6 weeks)

Graph Theory

  • Graph representation (adjacency matrix, adjacency list)
  • Graph traversal algorithms (BFS, DFS)
  • Shortest path problems
  • Minimum spanning trees
  • Graph connectivity and components

Computational Geometry

  • Point, line, and polygon representations
  • Convex hulls
  • Line intersection algorithms
  • Polygon triangulation
  • Voronoi diagrams and Delaunay triangulation
  • Visibility graphs
  • Computational complexity in geometry

Probability and Statistics

  • Probability distributions (Gaussian, uniform, exponential)
  • Bayesian probability
  • Random sampling methods
  • Monte Carlo methods
  • Markov chains and processes

Optimization

  • Linear and nonlinear programming
  • Gradient-based optimization
  • Heuristic search methods
  • Constrained optimization
  • Multi-objective optimization

Linear Algebra

  • Vector operations and transformations
  • Matrix operations
  • Eigenvalues and eigenvectors
  • Coordinate transformations (2D/3D)

Phase 2: Configuration Space and World Representation (4-5 weeks)

Configuration Space (C-Space)

  • Definition and properties
  • C-space for point robots
  • C-space for rigid bodies (SE(2), SE(3))
  • C-space obstacles (C-obstacles)
  • Forward and inverse obstacles
  • Degrees of freedom analysis

Environment Representation

  • Occupancy grids and maps
  • Quadtrees and octrees
  • Probabilistic occupancy maps
  • Cost maps and inflation layers
  • Multi-resolution representations
  • 3D voxel grids

Map Types

  • Metric maps
  • Topological maps
  • Hybrid maps
  • Semantic maps
  • Dynamic maps

Collision Detection

  • Bounding volume hierarchies (AABB, OBB, spheres)
  • Gilbert-Johnson-Keerthi (GJK) algorithm
  • Separating axis theorem
  • Distance computation
  • Continuous collision detection

Phase 3: Classical Path Planning Algorithms (6-8 weeks)

Graph Search Algorithms

  • Dijkstra's algorithm
  • Breadth-First Search (BFS)
  • Depth-First Search (DFS)
  • Uniform Cost Search
  • Bellman-Ford algorithm

Informed Search Algorithms

  • A* (A-star) algorithm
  • Weighted A*
  • Iterative Deepening A* (IDA*)
  • Greedy Best-First Search
  • Heuristic design (Euclidean, Manhattan, Octile)

A* Variants and Optimizations

  • Anytime A*
  • ARA* (Anytime Repairing A*)
  • D* (Dynamic A*)
  • D* Lite
  • Field D*
  • Theta* (any-angle path planning)
  • Jump Point Search (JPS)

Potential Field Methods

  • Attractive and repulsive potentials
  • Gradient descent navigation
  • Local minima problem
  • Potential field functions
  • Navigation functions

Cell Decomposition

  • Exact cell decomposition
  • Approximate cell decomposition
  • Trapezoidal decomposition
  • Boustrophedon decomposition
  • Morse decomposition

Roadmap Methods

  • Visibility graphs
  • Voronoi diagrams
  • Silhouette methods
  • Probabilistic roadmaps (basic concepts)

Phase 4: Sampling-Based Planning (6-8 weeks)

Probabilistic Roadmaps (PRM)

  • Construction phase and query phase
  • Sampling strategies
  • Connection strategies (k-nearest, radius)
  • Lazy PRM
  • PRM* (asymptotically optimal)

Rapidly-Exploring Random Trees (RRT)

  • Basic RRT algorithm
  • RRT-Connect (bidirectional RRT)
  • Goal-biased sampling
  • Informed RRT*
  • RRT* (optimal RRT)

RRT Variants

  • Kinodynamic RRT
  • RRT with dynamics
  • Anytime RRT
  • Execution Extended RRT (ERRT)
  • Transition-based RRT (T-RRT)
  • RRT-Sharp

Advanced Sampling Techniques

  • Quasi-random sampling (Halton, Sobol sequences)
  • Bridge test sampling
  • Gaussian sampling
  • Obstacle-based sampling
  • Adaptive sampling strategies

Batch Informed Trees

  • BIT* (Batch Informed Trees)
  • ABIT* (Advanced BIT)
  • FMT* (Fast Marching Tree)
  • Lower Bound Tree RRT (LBT-RRT)

Phase 5: Trajectory Optimization (5-7 weeks)

Parametric Curve Representation

  • Polynomial trajectories
  • Bezier curves
  • B-splines and NURBS
  • Clothoid/Euler spirals
  • Dubins paths
  • Reeds-Shepp curves

Smoothing and Post-Processing

  • Shortcut methods
  • Path simplification (Douglas-Peucker)
  • Elastic band methods
  • Gradient-based smoothing
  • B-spline fitting

Optimization-Based Planning

  • Trajectory optimization formulation
  • Shooting methods (direct and indirect)
  • Collocation methods
  • Sequential Quadratic Programming (SQP)
  • Interior point methods

Kinodynamic Planning

  • Differential constraints
  • State-space planning
  • Velocity and acceleration constraints
  • Control-based planning
  • Motion primitives

Optimal Control Methods

  • Minimum time trajectories
  • Minimum energy trajectories
  • Pontryagin's minimum principle
  • Hamilton-Jacobi-Bellman equation
  • Dynamic programming

Phase 6: Navigation for Specific Robot Types (6-8 weeks)

Differential Drive Robots

  • Kinematics and dynamics
  • Pure pursuit controller
  • Stanley controller
  • Vector field histograms (VFH)
  • Dynamic Window Approach (DWA)

Car-Like Robots (Ackermann Steering)

  • Bicycle model
  • Kinematic constraints
  • Dubins and Reeds-Shepp paths
  • Hybrid A* for parking
  • Lattice-based planning

Holonomic Robots

  • Omnidirectional motion
  • Mecanum wheels
  • Direct path planning
  • Time-optimal trajectories

Aerial Vehicles (UAVs/Drones)

  • 3D path planning
  • Altitude constraints
  • Wind considerations
  • Coverage path planning
  • Multi-rotor dynamics

Manipulator Path Planning

  • Joint space vs. task space
  • Cartesian path planning
  • Collision checking for links
  • Self-collision avoidance
  • Constrained motion planning

Legged Robots

  • Footstep planning
  • Terrain analysis
  • Contact planning
  • Balance constraints
  • Gait patterns

Phase 7: Dynamic and Real-Time Planning (5-7 weeks)

Reactive Navigation

  • Velocity obstacles (VO)
  • Reciprocal velocity obstacles (RVO)
  • Dynamic Window Approach (DWA)
  • Timed Elastic Band (TEB)
  • Model Predictive Control (MPC) for navigation

Replanning Strategies

  • Incremental replanning
  • Repair-based methods
  • Anytime algorithms
  • Real-time adaptive planning
  • Receding horizon planning

Moving Obstacles

  • Prediction of obstacle motion
  • Space-time planning
  • Velocity space planning
  • Safe interval path planning
  • Time-parameterized planning

Multi-Agent Path Finding (MAPF)

  • Centralized vs. decentralized planning
  • Conflict-Based Search (CBS)
  • Priority-based planning
  • Velocity obstacles for multiple agents
  • Coordination spaces

Phase 8: Localization and SLAM (6-8 weeks)

Localization

  • Markov localization
  • Kalman Filter localization
  • Extended Kalman Filter (EKF)
  • Particle Filter (Monte Carlo Localization)
  • Adaptive Monte Carlo Localization (AMCL)

Simultaneous Localization and Mapping (SLAM)

  • EKF-SLAM
  • FastSLAM (particle-based)
  • Graph-SLAM
  • Pose graph optimization
  • Bundle adjustment

Visual SLAM

  • Feature-based methods (ORB-SLAM, VINS)
  • Direct methods (LSD-SLAM, DSO)
  • Visual odometry
  • Loop closure detection
  • Place recognition

LiDAR-Based SLAM

  • ICP (Iterative Closest Point)
  • NDT (Normal Distributions Transform)
  • LOAM (LiDAR Odometry and Mapping)
  • Cartographer
  • LIO-SAM

Sensor Fusion

  • Multi-sensor integration
  • IMU integration
  • GPS/GNSS fusion
  • Camera-LiDAR fusion
  • Sensor calibration

Phase 9: Advanced Topics (8-10 weeks)

Learning-Based Planning

  • Neural network path planning
  • Deep Reinforcement Learning for navigation
  • Imitation learning
  • Value iteration networks
  • Graph neural networks for planning

Semantic Navigation

  • Object-goal navigation
  • Language-guided navigation
  • Semantic mapping
  • Scene understanding
  • Vision-language models for navigation

Coverage Path Planning

  • Area coverage algorithms
  • Sweeping patterns
  • Multi-robot coverage
  • Online coverage
  • 3D coverage (inspection tasks)

Uncertainty-Aware Planning

  • Belief space planning
  • POMDPs for navigation
  • Risk-aware planning
  • Chance-constrained planning
  • Information-theoretic planning

Exploration and Frontier-Based Planning

  • Frontier detection
  • Information gain
  • Next-best-view planning
  • Active SLAM
  • Exploration-exploitation trade-off

2. Major Algorithms, Techniques, and Tools

Core Path Planning Algorithms

Graph-Based

  • Dijkstra's Algorithm
  • A* and variants (Weighted A*, ARA*, AD*)
  • D* and D* Lite
  • Jump Point Search (JPS)
  • Theta* and Lazy Theta*
  • Field D*
  • LPA* (Lifelong Planning A*)

Sampling-Based

  • PRM (Probabilistic Roadmap)
  • PRM*
  • RRT (Rapidly-exploring Random Tree)
  • RRT-Connect
  • RRT*
  • Informed RRT*
  • BIT* (Batch Informed Trees)
  • FMT* (Fast Marching Tree)
  • SST (Stable Sparse RRT)

Optimization-Based

  • CHOMP (Covariant Hamiltonian Optimization for Motion Planning)
  • TrajOpt (Trajectory Optimization)
  • STOMP (Stochastic Trajectory Optimization)
  • GPMP (Gaussian Process Motion Planning)
  • iLQR/iLQG for trajectory optimization
  • Sequential Convex Programming

Reactive/Local Planning

  • Dynamic Window Approach (DWA)
  • Timed Elastic Band (TEB)
  • Velocity Obstacles (VO)
  • Reciprocal Velocity Obstacles (RVO/ORCA)
  • Vector Field Histogram (VFH/VFH+)
  • Nearness Diagram (ND)
  • Curvature Velocity Method

Kinodynamic Planning

  • State Space RRT
  • Kinodynamic RRT
  • Hybrid A*
  • LQR-RRT
  • SST (Stable Sparse Tree)
  • Motion Primitives/Lattice Planning

Multi-Agent Planning

  • Conflict-Based Search (CBS)
  • Enhanced CBS (ECBS)
  • M* (Multi-Agent A*)
  • ORCA (Optimal Reciprocal Collision Avoidance)
  • Push and Swap
  • Token Passing

SLAM Algorithms

2D SLAM

  • GMapping (Rao-Blackwellized Particle Filter)
  • Hector SLAM (scan matching)
  • Karto SLAM (graph-based)
  • Cartographer (Google)
  • CoreSLAM (lightweight)

3D SLAM

  • LOAM (LiDAR Odometry and Mapping)
  • LeGO-LOAM
  • LIO-SAM (LiDAR-Inertial)
  • FAST-LIO
  • HDL Graph SLAM

Visual SLAM

  • ORB-SLAM2/3
  • VINS-Mono/Fusion
  • LSD-SLAM
  • DSO (Direct Sparse Odometry)
  • RTAB-Map (RGB-D SLAM)
  • OpenVSLAM

Multi-Sensor SLAM

  • MSCKF (Multi-State Constraint Kalman Filter)
  • ROVIO (Robust Visual Inertial Odometry)
  • Kimera (semantic SLAM)
  • ElasticFusion

Software Tools and Libraries

Planning Libraries

  • OMPL (Open Motion Planning Library) - comprehensive sampling-based planning
  • MoveIt - ROS manipulation and planning
  • SBPL (Search-Based Planning Library) - lattice planners
  • MRPT (Mobile Robot Programming Toolkit)
  • FCL (Flexible Collision Library)

ROS/ROS2 Packages

  • Navigation Stack (move_base)
  • Nav2 (ROS2 navigation)
  • Global Planner (Dijkstra, A*)
  • Local Planner (DWA, TEB)
  • Costmap2D
  • AMCL (localization)

Simulation Environments

  • Gazebo
  • Stage/Stdr
  • CoppeliaSim (V-REP)
  • Webots
  • PyBullet
  • CARLA (autonomous driving)
  • AirSim (UAV/car simulation)
  • Isaac Sim (NVIDIA)

Mapping Tools

  • Cartographer
  • SLAM Toolbox
  • OctoMap (3D occupancy)
  • Elevation Mapping
  • GridMap library

Visualization

  • RViz/RViz2
  • Matplotlib
  • Plotly
  • Three.js for web visualization
  • Open3D (3D visualization)

Programming Languages & Frameworks

  • C++ (performance-critical code)
  • Python (rapid prototyping, learning)
  • ROS/ROS2
  • MATLAB/Simulink
  • Julia (emerging for robotics)

Path Smoothing Libraries

  • Cubic Splines
  • CasADi (optimization)
  • SciPy (Python optimization)
  • IPOPT, SNOPT (nonlinear optimization)

Specialized Tools

  • Autonomous Driving: Apollo (Baidu), Autoware, CARLA, LGSVL Simulator, OpenPilot
  • Drone Navigation: PX4 (flight stack), ArduPilot, RotorS (Gazebo), AirSim, Flightmare
  • Multi-Robot Systems: ROS Multi-Master, Fleet Management Systems, Swarm simulators

3. Cutting-Edge Developments (2023-2025)

Learning-Based Navigation

Foundation Models for Navigation

  • Vision-Language-Action models (VLA) for navigation
  • Large Language Models (LLMs) for high-level planning
  • Vision-Language Navigation (VLN) systems
  • Zero-shot navigation with pre-trained models
  • Multi-modal navigation systems

Neural Planners

  • Value Iteration Networks (VIN)
  • Differentiable planning networks
  • Graph Neural Networks (GNN) for planning
  • Neural Motion Planning (NMP)
  • Deep reinforcement learning (PPO, SAC) for navigation
  • World models for predictive planning

Imitation and Learning from Demonstration

  • Behavioral cloning for navigation
  • Inverse reinforcement learning
  • Goal-conditioned policies
  • Learning from human demonstrations
  • Sim-to-real transfer techniques

Semantic and Context-Aware Navigation

Semantic Understanding

  • Object-goal navigation ("find the chair")
  • Audio-visual navigation
  • Scene graph navigation
  • Affordance-based planning
  • Human-aware navigation (social navigation)

Embodied AI

  • Embodied Question Answering
  • Vision-and-Language Navigation (VLN)
  • Multi-object navigation
  • Instruction following
  • Interactive exploration

Robust and Safe Navigation

Safety and Verification

  • Formal verification of planning algorithms
  • Control Barrier Functions (CBF) for navigation
  • Reachability analysis
  • Hamilton-Jacobi reachability
  • Probabilistic safety guarantees

Uncertainty-Aware Methods

  • Risk-aware planning
  • Distributionally robust planning
  • Chance-constrained optimization
  • Conformal prediction for safety

Advanced SLAM and Perception

Neural SLAM

  • NeRF-based SLAM (Neural Radiance Fields)
  • 3D Gaussian Splatting for mapping
  • Implicit neural representations
  • Deep learning feature descriptors
  • End-to-end learned SLAM

Multi-Modal SLAM

  • Radar-camera-LiDAR fusion
  • Event camera SLAM
  • Thermal imaging integration
  • 4D radar SLAM

Dynamic Environments

  • Dynamic SLAM with moving objects
  • Panoptic mapping (stuff + things)
  • Temporal mapping
  • 4D occupancy mapping

Specialized Applications

Off-Road and Challenging Terrains

  • Learned terrain traversability
  • Self-supervised learning from experience
  • Proprioceptive navigation
  • Physics-informed planning

Long-Horizon Planning

  • Hierarchical planning with learning
  • Task and motion planning (TAMP)
  • Symbolic planning integration
  • Abstract state spaces

Swarm and Multi-Agent

  • Decentralized learning for swarms
  • Emergent behaviors
  • Large-scale coordination (100+ agents)
  • Communication-efficient algorithms

Edge Computing and On-Device

  • Efficient neural networks (quantization, pruning)
  • Neuromorphic navigation
  • Event-based processing
  • Real-time optimization on embedded systems

Novel Sensing Modalities

Unconventional Sensors

  • mmWave radar for navigation
  • Sonar and acoustic navigation
  • Magnetic field navigation
  • Bio-inspired sensing (whiskers, antennae)

Learned Representations

  • Self-supervised feature learning
  • Contrastive learning for place recognition
  • Metric learning for localization

4. Project Ideas by Level

Beginner Projects (1-3 weeks each)

1. Grid-Based A* Pathfinder

  • Implement A* on 2D occupancy grid
  • Visualize search process
  • Compare different heuristics
  • Add diagonal movement

2. Dijkstra vs A* Comparison

  • Implement both algorithms
  • Benchmark performance
  • Visualize nodes explored
  • Test on various maps

3. Potential Field Navigation

  • Simple attractive/repulsive fields
  • Navigate to goal avoiding obstacles
  • Identify and handle local minima
  • Visualization of field gradients

4. Pure Pursuit Path Tracker

  • Implement pure pursuit for car-like robot
  • Follow pre-defined path
  • Tune look-ahead distance
  • Simulate in 2D environment

5. Occupancy Grid Mapping

  • Build 2D map from simulated sensors
  • Implement probabilistic updates
  • Log-odds representation
  • Visualize in real-time

6. Dead Reckoning Navigator

  • Odometry-based navigation
  • Track position errors over time
  • Visualize drift
  • Compare with ground truth

Intermediate Projects (3-6 weeks each)

7. RRT Path Planner Implementation

  • Basic RRT from scratch
  • Goal-biased sampling
  • RRT-Connect bidirectional search
  • Compare performance metrics

8. Probabilistic Roadmap (PRM)

  • Build roadmap in 2D/3D space
  • Implement sampling strategies
  • Connection phase optimization
  • Query phase for multiple goals

9. Dynamic Window Approach (DWA)

  • Local planning for differential drive
  • Velocity space sampling
  • Cost function design
  • Real-time obstacle avoidance

10. Hybrid A* for Car Parking

  • Implement Hybrid A* algorithm
  • Include Reeds-Shepp heuristics
  • Handle kinematic constraints
  • Simulate parallel parking scenario

11. Monte Carlo Localization (MCL)

  • Particle filter implementation
  • Sensor models (range, bearing)
  • Resampling strategies
  • Kidnapped robot problem

12. Timed Elastic Band (TEB) Planner

  • Implement TEB local planner
  • Optimization-based trajectory generation
  • Dynamic obstacle avoidance
  • Test with various robot models

13. Coverage Path Planning

  • Boustrophedon decomposition
  • Complete area coverage
  • Optimize for minimal turns
  • Handle complex shaped areas

14. Jump Point Search on Grid

  • Implement JPS algorithm
  • Compare with standard A*
  • Benchmark on large maps
  • Visualize jump points

Advanced Projects (6-10 weeks each)

15. RRT* with Informed Sampling

  • Implement asymptotically optimal RRT*
  • Add informed set sampling
  • Compare convergence rates
  • 3D implementation

16. Full 2D SLAM System

  • EKF-SLAM or GraphSLAM
  • Landmark detection and association
  • Loop closure detection
  • Real-time mapping and localization

17. Visual Odometry Pipeline

  • Feature extraction (ORB, SIFT)
  • Feature matching and tracking
  • Essential matrix estimation
  • Scale estimation with IMU

18. Multi-Robot Path Planning

  • Implement CBS or ORCA
  • Coordinate multiple agents
  • Deadlock detection and resolution
  • Scalability testing

19. Trajectory Optimization with CHOMP

  • Implement CHOMP algorithm
  • Gradient-based optimization
  • Collision cost formulation
  • Compare with sampling-based methods

20. LiDAR-Based SLAM (2D)

  • Scan matching (ICP or NDT)
  • Pose graph construction
  • Loop closure with scan matching
  • Real sensor data processing

21. Semantic Navigation System

  • Object detection integration
  • Semantic map building
  • Goal specification by object class
  • Navigate to semantic goals

22. MPC for Path Following

  • Model Predictive Control formulation
  • Receding horizon planning
  • Constraint handling
  • Real-time implementation

23. Replanning with D* Lite

  • Implement D* Lite algorithm
  • Dynamic environment changes
  • Incremental replanning
  • Compare with repeated A*

24. ROS Navigation Stack Integration

  • Configure move_base
  • Custom costmap layers
  • Custom local/global planners
  • Deploy on physical robot

Expert/Research Projects (10+ weeks)

25. Learning-Based Navigation with DRL

  • Train RL agent (PPO, SAC) for navigation
  • Curriculum learning approach
  • Sim-to-real transfer
  • Compare with classical methods

26. Visual-Inertial SLAM

  • Implement VINS or similar
  • Tight sensor fusion
  • Initialization procedures
  • Real-world testing

27. Multi-Floor 3D Navigation

  • 3D path planning across levels
  • Elevator/stair navigation
  • 3D semantic mapping
  • Long-term autonomy

28. Social Navigation in Crowds

  • Human trajectory prediction
  • Socially-aware cost functions
  • Implement Social Force Model
  • Real-world deployment

29. Neural Motion Planning

  • Train neural network planner
  • Value Iteration Networks or similar
  • Generalization to new environments
  • Comparison with OMPL

30. Autonomous Exploration System

  • Frontier-based exploration
  • Information-theoretic planning
  • Active SLAM
  • Complete unknown environment

31. Vision-Language Navigation

  • Implement VLN system
  • Natural language instruction following
  • Vision transformer integration
  • Test on ALFRED or similar benchmarks

32. BIT* Implementation and Analysis

  • Implement Batch Informed Trees
  • Performance analysis
  • Comparison with RRT* and FMT*
  • High-dimensional spaces

33. NeRF-SLAM

  • Neural radiance field mapping
  • Real-time reconstruction
  • Pose estimation from NeRF
  • Dense 3D mapping

34. Cooperative Multi-Agent Coverage

  • Distributed coverage algorithm
  • Communication protocols
  • Voronoi-based partitioning
  • Real multi-robot deployment

35. End-to-End Learning for Navigation

  • Visuomotor policies
  • Direct sensor-to-action mapping
  • Dataset collection
  • Deployment and safety analysis

36. Manipulation with Mobile Base

  • Mobile manipulation planning
  • Coordinate arm and base
  • Whole-body planning
  • Pick and place in clutter

37. Terrain-Aware Planning for Legged Robots

  • Footstep planning
  • Terrain classification
  • Contact planning
  • Dynamic locomotion

38. Urban Autonomous Navigation

  • HD map integration
  • Traffic rule compliance
  • Behavior prediction
  • Deploy in simulator (CARLA)

39. Uncertainty-Aware Planning with POMDPs

  • Belief space planning
  • Online POMDP solver
  • Information gathering behaviors
  • Comparison with deterministic planning

40. Quantum-Inspired Path Planning

  • Quantum annealing for optimization
  • Hybrid classical-quantum algorithms
  • Comparative analysis
  • Scalability study

Recommended Learning Strategy

Phase-by-Phase Approach

Months 1-2: Foundations

  • Focus on mathematical prerequisites
  • Implement basic graph algorithms
  • Learn C++ or Python thoroughly
  • Get comfortable with ROS basics

Months 3-4: Classical Algorithms

  • Implement A* from scratch
  • Study grid-based methods
  • Build intuition with visualizations
  • Complete beginner projects

Months 5-6: Sampling-Based Methods

  • Deep dive into RRT family
  • Implement PRM and variants
  • Understand probabilistic completeness
  • Work on intermediate projects

Months 7-8: Real Robot Systems

  • Learn ROS navigation stack
  • Work with real sensors (LiDAR, cameras)
  • Implement localization
  • Deploy on hardware

Months 9-10: Advanced Topics

  • Study SLAM algorithms
  • Trajectory optimization
  • Multi-agent systems
  • Advanced projects

Months 11-12: Specialization

  • Choose focus area (learning, SLAM, multi-agent, etc.)
  • Research current literature
  • Expert projects
  • Contribute to open-source

Essential Resources

Textbooks

  • "Planning Algorithms" by Steven LaValle (free online)
  • "Probabilistic Robotics" by Thrun, Burgard, Fox
  • "Principles of Robot Motion" by Choset et al.
  • "Robotics, Vision and Control" by Peter Corke

Online Courses

  • Coursera: "Motion Planning for Self-Driving Cars"
  • Coursera: "Robotics: Computational Motion Planning"
  • MIT OCW: "Underactuated Robotics"
  • ETH Zurich: "Autonomous Mobile Robots"

Research Venues

  • ICRA, IROS (robotics conferences)
  • RSS (Robotics: Science and Systems)
  • CoRL (Conference on Robot Learning)
  • ArXiv (cs.RO category)

Practice Platforms

  • ROS/ROS2 tutorials
  • OMPL documentation and demos
  • GitHub repositories of SLAM systems
  • Robotics competitions

This comprehensive roadmap provides a 12-month learning journey with flexibility to adjust based on your interests and career goals. The key is to balance theory with hands-on implementation throughout your learning process.