Dec 16, 2025 • Entry #004

Project Kickoff: Autonomous Forklift 🚜

New industrial challenge unlocked. Designing a navigation stack for automated pallet moving using ROS 2 and LiDAR.

The Objective

The goal is to transform a standard warehouse forklift into an Autonomous Mobile Robot (AMR). It needs to:

Map: Navigate a warehouse environment autonomously.
Detect: Identify pallets using computer vision.
Manipulate: Approach and lift loads with precision (Visual Servoing).

The Stack

This will be a heavy ROS 2 (Jazzy) project.

Navigation: Nav2 Stack for path planning and obstacle avoidance.
Localization: SLAM Toolbox (initially) + AMCL.
Perception: 2D LiDAR for mapping + Depth Camera (Realsense) for pallet pose estimation.

$/ perception/pallet_detector.py

# TODO: Implement PnP (Perspective-n-Point) for pallet pose
class PalletDetector(Node):
    def __init__(self):
        super().__init__('pallet_detector')
        self.create_subscription(Image, '/camera/depth', self.depth_callback, 10)

Current Status

Setting up the URDF model for Gazebo simulation to test the ackermann steering kinematics before touching real hardware. Safety first!