An object detection system designed to enhance astronaut safety during EVAs by identifying critical mission equipment through cutting-edge AI.
spacevision_ai is an intelligent object detection system powered by YOLOv8 that helps astronauts identify vital tools, such as fire extinguishers, oxygen tanks, and toolboxes, in real-time. (Future Vision) Integrated as a vision assistant into space helmets, it acts as a second pair of eyes, improving spatial awareness and reaction time during spacewalks and repairs.
In the vast void of space, visibility, response time, and situational awareness can mean the difference between life and death. This system ensures astronauts:
- Don’t miss critical equipment
- Can work more independently with AI assistance
- Get real-time alerts in harsh, low-visibility environments
| Component | Technology |
|---|---|
| AI Model | YOLOv8 (Ultralytics) |
| Programming Lang | Python |
| Hardware | NVIDIA RTX 4060 (8GB VRAM) |
| IDE | VS Code |
- 📦 Identifies fire extinguishers, oxygen tanks, and toolboxes
- 🧠 Trained on a custom dataset tailored for space utility
- 📊 Achieved mAP@50: 96.4%
- 💡 Designed with future integration in space helmet AR HUDs
| Metric | Value |
|---|---|
| mAP@50 | 96.4% |
| Precision (mean) | 99.31% |
| Recall (mean) | 92.38% |