Brilliant Labs' Monocle Is an Open Source Augmented Reality Wearable for the "Imaginative Hacker"

Brilliant Labs is aiming to lower the barrier to entry for compact, wearable computer vision and augmented reality projects with its Monocle — an open source gadget with display, camera, and on-board processing smarts you can simply hold up to your eye or clip to your glasses.

"Monocle is a tiny heads-up display which clips onto your existing eyewear," Brilliant Labs explains of its creation. "It's packed full of powerful hardware which is perfect for when you're on the go. It connects to your mobile phone over Bluetooth, and includes a few handy sensors such as touch buttons, camera, and microphone. The included FPGA is perfect for computer vision, AI, or graphics acceleration directly on the device. Monocle runs open source software and provides an easy way to get started using MicroPython."

The gadget, shaped like an old-fashioned monocle but considerably thicker, is built around a Nordic Semi nRF52832 system-ion-chip (SOC) with a separate GoWin GW1NR-9 field-programmable gate array (FPGA) connected over an SPI bus. The FPGA handles the display, a 640×400 Sony ECX336CN micro-OLED display with the optics required to work as a near-eye augmented reality display, and there's an Omnivision OV5640 camera sensor and a TDK ICS-41351 microphone. A capacitive touch sensor sits to the top of the enclosure, and a clip fastens the device to your glasses.

Power is provided using a 70mA internal battery — chosen to keep the overall weight down to a svelte 15g — which provides a claimed two-hour runtime, and is automatically refreshed up to six times using a larger battery in the bundled charging case. Out-of-the-box, the device comes running a MicroPython port with capabilities including 16× "super-zoom" vision, instant replay with slow-motion mode, and smartphone-connected still image and video capture and sharing.

The company is hoping to see developers take the technology and run with it: there's room on the FPGA to run on-device machine learning models, with the company highlighting its capabilities for object detection and image segmentation as well as for providing head-up telemetry for external hardware over its Bluetooth connection. Brilliant Labs is also promoting the Monocle's compatibility with Streamlogic's drag-and-drop computer vision development environment, to ease newcomers into taking advantage of its capabilities.

Brilliant Labs is now taking orders for the Monocle at $349, with the first batch due to begin shipping this month. The project source code, meanwhile, is available on GitHub under the permissive MIT license.