🚀 New from Groundlight: Object Counting 🔢 We’re excited to announce a major step forward in simplifying computer vision — the launch of our new Counting Mode, for accurate, real-time object counting. Whether you're counting cars 🏎️ , people 👯, or pallets, our latest

🚨Today’s the Day! Live at 12 PM PST 🚨 Learn how the powerful combination of Groundlight and @Universal_Robots automates machine tending - you're not going to want to miss this 45-minute webinar!

🚀 Live at Automate Show! Catch our very own Tim Huff at the Orbbec booth (#9004) today and tomorrow morning, where he’s demoing the Groundlight machine vision integration with Universal Robots. Don’t miss this chance to see the future of robotics up close. 👀🤖

If you missed out on last week's webinar that featured a powerful combination of Universal Robots cobots and Groundlight, here's a snippet of Groundlight's Tim Huff and Julia Gall leading a demo - from checking if the CNC door is open to checking that the part is properly

We made an open-source MCP server github.com/groundlight/mc… that turns HuggingFace zero-shot object detection pipelines into tools that Claude and others can use to locate objects or zoom (crop) to an object. Conceptually vision capabilities as tools are complementary to VLM's

Groundlight is now available for use with Agentic AI as the Groundlight MCP Server. You can build a sophisticated video stream monitoring computer vision application complete with alerts by having a conversation with Claude Desktop or your favorite AI coding agent.

Create a computer vision model to count objects using simple natural language prompts (ex: "Label each package in the image"). A sneak peak from Groundlight's dashboard: