UCHub

Camera to be operated while plugged in to a raspberry pi
ESP32 on a Robot Arm, enabling the MCP server to easily detect the device

Inspiration

MCP is one of the hottest topics right now. It allows AI models to talk to software and do amazing things. But hardware developers got left behind in this. That’s what inspired us to build UCHub — to bring the same simplicity and power of MCP to the physical world, so hardware developers can finally join the conversation.

What it does

It standardizes how hardware interfaces and communicates extending the contextual understanding of MCP into the physical world. With UCHub, AI agents can easily interact with hardware devices through a unified communication layer. Despite how messy the setup or how many different physical connecting method is used, USB, Wi-Fi, Bluetooth, or GPIO, UCHub makes everything work together effortlessly.

How we built it

Read usb specification and usip protocol documentation, we were able to build a framework that can work on almost any microcontroller. Then we built a MCP service that uses this feature to reach to hardwares.

Challenges we ran into

We did not have any prior experience with MCP. So, developing that part took us a lot of time.

Accomplishments that we're proud of

One of our proudest moments was when we successfully used Claude Desktop to directly interact with a device remotely connected through an ESP32. Seeing an AI agent communicate with real hardware in real time proved that our framework works. We felt how this could help hardware devlopers in many aspects.