Nav2 — Path Planning for Bipedal Humanoid Movement

What is Nav2?

Nav2 (Navigation 2) is the standard navigation framework for ROS 2. It provides:

• Costmaps — 2D/3D representations of obstacles
• Path Planners — algorithms to find a safe path from A to B
• Controllers — algorithms to follow the planned path
• Behavior Trees — high-level task management

The Nav2 Stack

Goal Pose
    │
    ▼
┌─────────────────┐
│   BT Navigator  │  ← Behavior Tree manager
└────────┬────────┘
         │
    ┌────▼────┐
    │ Planner │  ← Global path planning (NavFn, Smac)
    └────┬────┘
         │
    ┌────▼────┐
    │Controller│  ← Local trajectory following (DWB, RPP)
    └────┬────┘
         │
    ┌────▼────┐
    │ Costmap │  ← Obstacle representation from sensors
    └─────────┘
         │
         ▼
   /cmd_vel → Robot

Launching Nav2

# Basic Nav2 launch with Gazebo
ros2 launch nav2_bringup tb3_simulation_launch.py

# With your custom robot
ros2 launch nav2_bringup bringup_launch.py \
    map:=/path/to/map.yaml \
    params_file:=/path/to/nav2_params.yaml

Sending Navigation Goals from Python

from geometry_msgs.msg import PoseStamped
from nav2_simple_commander.robot_navigator import BasicNavigator
import rclpy

rclpy.init()
navigator = BasicNavigator()

# Wait for Nav2 to be ready
navigator.waitUntilNav2Active()

# Define goal pose
goal_pose = PoseStamped()
goal_pose.header.frame_id = 'map'
goal_pose.header.stamp = navigator.get_clock().now().to_msg()
goal_pose.pose.position.x = 2.0
goal_pose.pose.position.y = 1.5
goal_pose.pose.orientation.w = 1.0  # Facing forward

# Send goal
navigator.goToPose(goal_pose)

# Wait for completion
while not navigator.isTaskComplete():
    feedback = navigator.getFeedback()
    print(f'Distance remaining: {feedback.distance_remaining:.2f}m')

result = navigator.getResult()
print(f'Navigation result: {result}')
rclpy.shutdown()

Bipedal Considerations

Wheeled robots and bipedal humanoids have different navigation challenges:

Challenge	Wheeled Robot	Bipedal Humanoid
Stability	Always stable	Must maintain balance
Footstep planning	Not needed	Required
Terrain	Flat surfaces	Stairs, uneven ground
Recovery	Simple	Complex (fall recovery)

For bipedal robots, you need to add a footstep planner on top of Nav2. Research projects use:

• Humanoid Path Planner (HPP)
• Footstep Planning in IHMC
• MIT Footstep Planner

---

Module 3 Complete! Move on to Module 4: Vision-Language-Action →