Visual Intelligence for ROS2 Robotics

Empower your robots with Groundlight's advanced visual question answering. Seamlessly integrate AI-driven perception into ROS2 projects, enabling natural language queries and real-time decision-making for smarter, more adaptable robotic systems.

Seamless Integration of AI Vision in ROS2

Incorporate powerful computer vision into your robotics projects with minimal effort.

Natural Language Queries

Ask questions about the robot's surroundings using plain English.

Real-time Processing

Analyze live video streams on the edge for immediate robot decision-making and actions.

Fully-managed Models

Groundlight trains a custom model for every task, every robot, learning exactly your processes and standards.

Easy Implementation for ROS2 Developers

The groundlight_ros package seamlessly integrates with existing ROS2 nodes and systems. Leverage our intuitive API and comprehensive documentation to quickly add advanced computer vision capabilities to your robotics projects.

Flexible Configuration

Customize image topics, question prompts, and other parameters effortlessly.

Human-in-the-Loop

Leverage 24/7 human oversight for reliable answers in uncertain situations.

Fast Edge Computing

Utilize on-device GPU for fast, on-robot inference capabilities.

Powering Next-Gen Robotics Applications

Enable more intelligent and adaptable robotic systems with visual understanding.

Anomaly Detection

Groundlight’s AI software is trained by your operators, learning and improving over time.

Continuous Improvement

Benefit from Groundlight's constantly evolving visual AI models and capabilities.

Versatile Applications

From remote inspection to logistics, or specialized manufacturing, unlock new robotics possibilities.

Install the Groundlight ROS package today

cd ~/ros2_ws/src
git clone git@github.com:groundlight/groundlight_ros.git
colcon build