Vienna – October 14, 2026 -- Viewpointsystem has launched an eye-tracking module that delivers up to 15 times longer battery runtime than conventional processing on a connected device, according to the company's performance measurements. The Digital Iris Module runs the complete image-based eye-tracking pipeline at 140 Hz on an ultra-low-power Ensemble microcontroller (MCU) from Alif Semiconductor, inside the glasses frame.
Power consumption, heat generation and limited space have so far made eye tracking hard to add to lightweight smart glasses, the company said. Conventional systems send images from eye-facing cameras to a main application processor in the glasses or a connected device. Processing those streams draws power, generates heat and adds system complexity.
Module moves the entire processing pipeline onto the MCU
The module pairs two eye-facing cameras with the MCU in a unit designed to fit into the bridge of the glasses frame. The MCU handles image preprocessing, gaze calculation accelerated by its integrated neural processing unit (NPU), and real-time detection of blinks, fixations and saccades. It outputs gaze data rather than camera images.
Target uses include real-time attention and focus tracking, as well as applications assessing brain health, cognitive changes and neurological function.
OEMs can drop the high-end application processor
Smart-glasses manufacturers no longer need a high-end application processor for eye tracking. The module stays compatible with existing application processors. The programmable MCU can also run additional AI-based functions on its NPU.
The MCU reaches 140 Hz and, depending on customer requirements and use cases, could reach up to 400 Hz. It also acquires and fuses data from an ambient light sensor, magnetometer and accelerometer on the edge. That frees the main processor for other tasks or low-power states.
Only gaze data leaves the chip
Sensitive eye-camera images stay on the MCU, and only gaze data is transmitted. Viewpointsystem described the hardware-and-software module as ready to integrate and adaptable to different smart-glasses designs, which it says reduces OEM development effort.
Nils Berger, CEO and founder of Viewpointsystem, said moving the whole pipeline to the edge lets OEMs integrate eye tracking across a wider range of designs, with or without a high-end application processor.
Alif positions Ensemble family for thermal-constrained eyewear
Mark Rootz, VP Global Marketing at Alif Semiconductor, said the implementation shows the Ensemble MCU can handle demanding real-time vision processing within the power and space limits of smart glasses, without the thermal overhead of an application processor.
Viewpointsystem calls the solution patented. It targets lightweight AI and AR glasses, display-less eyewear and specialized industrial applications where power, space and processing constraints have ruled out eye tracking.