# Sensor Fusion

> Sensor Fusion is an Agent Skill for development published in the a5c-ai/babysitter repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Multi-sensor fusion algorithms for perception in autonomous driving

- Skill name: `sensor-fusion`
- Category: [Development](https://www.openparable.com/skill/development/1)
- Source: https://github.com/a5c-ai/babysitter/blob/HEAD/library/specializations/domains/science/automotive-engineering/skills/sensor-fusion/SKILL.md
- Repository: a5c-ai/babysitter (1,548 GitHub stars when collected)

## SKILL.md

# Sensor Fusion Skill

## Purpose
Enable multi-sensor fusion algorithm development for autonomous driving perception including object detection, tracking, and environmental modeling.

## Capabilities
- Camera, radar, lidar data preprocessing
- Object detection fusion algorithms
- Tracking filter implementation (Kalman, EKF, UKF)
- Association algorithms (Hungarian, GNN, JPDA)
- Occupancy grid fusion
- Confidence estimation and sensor weighting
- Time synchronization handling
- Ground truth comparison and metrics

## Usage Guidelines
- Preprocess sensor data for consistent coordinate frames
- Select appropriate tracking filters based on object dynamics
- Implement robust association for multi-target scenarios
- Fuse sensor confidence for reliable perception
- Handle time delays and synchronization issues
- Validate fusion against ground truth data

## Dependencies
- ROS/ROS2
- TensorFlow
- PyTorch
- NVIDIA DriveWorks

## Process Integration
- ADA-001: Perception System Development
- ADA-002: Path Planning and Motion Control
- ADA-003: ADAS Feature Development
- ADA-004: Simulation and Virtual Validation

---

Source: https://www.openparable.com/skill/development/sensor-fusion (OpenParable). Cite the canonical URL above.
