Genesis is a semi-autonomous rover platform designed for outdoor exploration, GPS-guided waypoint navigation, and obstacle avoidance. The system operates using a hybrid architecture: high-level computer vision and decision-making are handled by an Nvidia Jetson Nano, while real-time motor control and sensor integration are managed by an Arduino Mega.
The platform features a custom Base Station for telemetry and control, as well as voice-command capabilities for hands-free operation.
- System Architecture
- Project Structure
- Key Features
- Hardware Specifications
- Getting Started
- Usage
- Documentation
- Contributors
The software is divided into two primary nodes communicating via radio telemetry:
- Onboard System: Runs on the rover. Handles autonomous navigation logic, depth-camera processing, and hardware actuation.
- Base Station: Runs on a host laptop. Provides a GUI for manual control, mode switching (Auto/Manual), and live status monitoring.
This repository is organized as follows:
base_station/: Ground control software.gui/: Contains Tkinter interface logic and configuration.run_station.py: Entry point for the laptop control application.
onboard/: The primary rover software (Jetson Nano).controls/: Algorithms for GPS navigation (gps.py) and autonomous logic (fullauto.py).handlers/: Interfaces for Serial communication (mega.py) and Manual control mapping.modules/: Drivers for the RealSense camera (camera.py).main.py: Entry point for the rover.
extra/: Experimental scripts and isolated tests.detection/: Standalone scripts for testing obstacle avoidance algorithms.imu/: Visualization tools for sensor data.
pcb/: Hardware design files.- Contains schematics and Gerber files for the custom power distribution board.
docs/: Detailed documentation and media assets.
- GPS Guided Navigation: autonomous traversal to waypoints using the Haversine formula and magnetometer heading correction.
- Vision-Based Avoidance: Uses an Intel RealSense D415 to detect obstacles. The pipeline utilizes Canny Edge Detection and depth thresholding to identify hazards.
- Voice Interaction: Supports voice commands (e.g., "Stop", "Go") and Q&A using NLP models.
- Telemetry: Long-range 433MHz radio link for status updates and manual override.
- Custom Electronics: Features a dedicated PCB for efficient power management between the computing units and high-torque motors.
- Compute: Nvidia Jetson Nano (4GB) & Arduino Mega 2560
- Vision: Intel RealSense D415 Depth Camera
- Sensors: Ublox Neo M8N GPS, HMC5883L Magnetometer
- Actuation: BTS7960 43A Motor Drivers
- Communication: 3DR 433MHz Telemetry Radio
- Power: 4S LiPo Battery (14.8V)
For a detailed breakdown of the electronics, refer to docs/hardware.md.
- Hardware: A fully assembled Genesis rover and a host laptop.
- Operating System: Ubuntu 18.04 (Jetson Nano), Linux/Windows/MacOS (Base Station).
- Python: Version 3.6 or higher.
On your control laptop, connect the telemetry radio and run:
cd base_station
python3 run_station.pySSH into the Jetson Nano or connect via terminal. Ensure the Arduino is connected via USB. Run:
cd onboard
python3 main.py- Manual: Use the WASD keys on the Base Station to drive the rover.
- Auto: Upload GPS coordinates via the GUI. The rover will calculate the bearing and distance, navigating autonomously while avoiding obstacles.
For deep dives into the specific algorithms and hardware designs, please refer to the docs/ folder:
- Software Manual: Detailed explanation of the navigation math (Haversine), logic flows, and Vision pipeline.
- Hardware Guide: Wiring diagrams and component specifics.
- PCB Schematics: Circuit diagrams for the custom board.
- Aiden Ross D'souza
- Telukunta Phani Raj
- Koyaneni Yashwant
- Shubham Pandere
- Yuvraj Gupta