By: Thomas Kim, Tharun Nayar, Carson Perkins, Kevin Garcia Varon
An autonomous mobile robot designed to navigate a warehouse environment with obstacle avoidance, task monitoring, and manual control through a Node-RED dashboard interface.
Built using the SCUTTLE robotics platform and powered by a Raspberry Pi running RPiOS, this project combines computer vision (OpenCV), LiDAR sensing, and local file I/O communication to enable flexible, modular control.
System-level overview of the SCUTTLE warehouse robot
The robot autonomously navigates a mapped indoor area using LiDAR and ultrasonic sensing for obstacle avoidance, while visual data from a front-mounted camera supports lane and object detection.
A Node-RED dashboard allows operators to view robot telemetry, task progress, and send simple control commands.
- Develop a functional autonomous navigation system for a warehouse-like environment.
- Implement real-time obstacle avoidance using LiDAR and ultrasonic sensors.
- Create a Node-RED dashboard for control, logging, and monitoring.
- Use Python for onboard logic, computer vision (OpenCV), and data exchange through local file I/O (no MQTT dependency).
- Integrate an intuitive UI for demonstration and educational purposes.
Software and communication architecture for SCUTTLE warehouse robot
| Layer | Description |
|---|---|
| Hardware | SCUTTLE base, Raspberry Pi 4/5, LiDAR sensor, Ultrasonic array, USB camera, motor controllers |
| Firmware | Motor drivers and low-level motion control |
| Software Stack | Python-based control and sensing nodes, OpenCV for vision processing, Node-RED dashboard UI |
| Communication | Local file I/O (JSON and CSV data exchange between Node-RED and Python) |
| User Interface | Web-accessible dashboard hosted by Node-RED for monitoring and control |
- LiDAR sensor continuously maps nearby objects within a predefined radius.
- Ultrasonic sensors act as redundancy for close-range detection.
- OpenCV processes camera input for path following, colored marker detection, or boundary recognition.
- Autonomous logic in Python integrates these data streams to make directional decisions and speed adjustments in real-time.
Concept mockup of Node-RED dashboard interface
The dashboard allows for:
- Manual control (forward, reverse, turn)
- Live telemetry (speed, battery, distance sensors, CPU temperature)
- Log viewing (mission progress, error logs)
- Mode switching between Autonomous and Manual
Data from Python scripts is stored as JSON/CSV in /home/pi/scuttle_data/, which Node-RED reads periodically to update the dashboard in real-time.
| Category | Tools / Libraries |
|---|---|
| Core Hardware | SCUTTLE Robot, Raspberry Pi 4/5 |
| Programming | Python 3, Node-RED |
| Computer Vision | OpenCV |
| Sensing | LiDAR (RPLidar A1/A2), HC-SR04 Ultrasonic Sensors |
| Data Handling | Local File I/O (JSON, CSV) |
| Operating System | Raspberry Pi OS (Debian-based) |
| Networking | Configured via wpa_supplicant.conf for Wi-Fi connectivity and remote access |
git clone https://github.com/yourusername/scuttle-warehouse-robot.git
cd scuttle-warehouse-robot