The Concept
Automating warehouse logistics requires precise navigation and manipulation. This project simulates a complete autonomous forklift system capable of transporting pallets between shelves.
It features a Mission Control GUI, a Waypoint Follower for graph-based navigation, and a custom Lift Controller for pallet manipulation.
System Architecture
Interface Node (GUI)
A Tkinter-based GUI for mission control. It publishes ROS 2 topics to trigger navigation and manipulation tasks.
Waypoint Follower (Nav2)
Uses BFS algorithm on a topological graph to plan paths. Sends velocity commands (`/cmd_vel`) to the robot via Nav2 stack (AMCL + Planner).
Lift Controller
Manages the pallet engagement. It detects the nearest pallet (< 3m) and "attaches" it to the forklift in the simulation, updating its position at 50Hz.
Mission Phases
- 01. NAV → ORIGIN: Go to pickup shelf
- 02. APPROACH: Blind approach (0.5 m/s)
- 03. PICKUP: Engage pallet
- 04. REVERSE: Back off safely
- 05. NAV → DEST: Go to dropoff shelf
- 06. DEPOSIT: Release pallet
- 07. NAV → HOME: Return to base
Graph Editor
Includes a custom web-based Graph Editor to design the warehouse topology. Export nodes and edges to GeoJSON for the Waypoint Follower.
Project Status
Fully Functional
- Full Mission Loop: From pickup to delivery and return home.
- Graph Navigation: Robust topological navigation using BFS.
Future Work
The current system is simulation-only. Next steps involve deploying the stack to a physical robot with real LIDARs and motor drivers.
Project Team
Project Context
Academic project for the University of Alicante. Exploring autonomous logistics solutions with ROS 2.