AI Meets Steampunk in Mind-Control Helmet

Wearable electronics usually don't go much beyond flashy LEDs or a small display, but DeepSea Developments decided to take things much further for the latest version of its Mind-Control Helmet. Built for Burning Man, the steampunk-inspired creation combines a Raspberry Pi, a local large language model, a remotely operated animatronic eye, live video streaming, and a custom web interface into a self-contained system that doesn't require an internet connection to function.

At the core of the helmet is a Raspberry Pi 5 that serves as the primary controller. Rather than simply running a few lighting effects, it hosts an entire software stack that creates a local Wi-Fi hotspot called "Mind Control." Anyone connected to that network can open a browser-based control panel to access the helmet's camera feed, adjust lighting effects, control the animatronic eye, communicate through the onboard speakers, or interact with the built-in AI assistant.

To make this possible, the Raspberry Pi is paired with a Raspberry Pi AI HAT+, providing dedicated neural processing hardware. That allows the helmet to run the 1.5-billion-parameter Qwen 2.5 language model directly on the device instead of relying on cloud services. Voice interaction is handled entirely onboard as well, with OpenWakeWord detecting spoken activation phrases, local speech-to-text software converting speech into prompts, and Piper generating spoken responses through the helmet's speakers.

A separate Seeed Studio XIAO ESP32S3 handles the jobs that benefit from a dedicated microcontroller. It drives the addressable LEDs, controls the servos that animate the eye, and receives commands from a handheld wireless controller. Separating those real-time tasks from the Raspberry Pi keeps the physical controls responsive while the main computer handles AI inference, networking, and video streaming.

One of the helmet's most noticeable features is its large mechanical eye. Built from an open-source 3D-printable design, the eye is finished with textured paint to match the rest of the steampunk aesthetic. Multiple servos give it the ability to pan, tilt, blink, and open its eyelids wide, and software can replay previously recorded movements to keep the eye wandering around naturally whenever nobody is actively controlling it.

Addressable LED strips and rings illuminate the goggles and surrounding frame with patterns ranging from Matrix-style animations to rainbow, comet, glitter, and chaos lighting effects. A front-mounted display replaces the static elements found on the original helmet with animated graphics including plasma, fire, and falling digital rain.

To help the remote operator see what the wearer sees, the project uses a Raspberry Pi Camera Module 3 NoIR mounted near the front of the helmet. The NoIR version was selected for its improved low-light performance, making it better suited for nighttime events. Audio travels in both directions through the web interface, allowing operators to hear the helmet's surroundings and speak through the integrated speakers.

While the finished helmet certainly looks at home in a steampunk setting, the project is really a showcase for modern embedded hardware. It combines AI acceleration, wireless networking, animatronics, computer vision, and custom electronics into a single wearable platform, all while operating entirely on its own. Whether it's used to entertain crowds at Burning Man or as a testbed for future ideas, the Mind-Control Helmet V2 shows just how capable today's embedded hardware platforms have become.