HKUST Bionic Eye Adds Full-Color Imaging

The hemispherical artificial retina pairs high pixel density with event-driven motion detection, according to a study in Nature Materials.

Updated on Nov. 13, 2026 in Robotics

Bold vector editorial illustration of a curved dome covered in colored pixel dots resting on a bench, evoking an artificial retina.
Researchers at the Hong Kong University of Science and Technology published a bionic eye with full-color imaging and a field of view above 160 degrees in Nature Materials. AI Illustration. Upload story photo >

Researchers at The Hong Kong University of Science and Technology have developed a high-resolution bionic eye with full-color imaging and a field of view exceeding 160 degrees. The system detects motion within the sensor itself, reporting 98.6% motion recognition accuracy in the study, and the findings were published in Nature Materials.

Why it matters

By processing motion inside the sensor rather than transmitting full image sequences, the design reduces data bandwidth demand by more than 99.95% compared with frame-based imaging under the study's test conditions. This points toward artificial vision systems that are lighter on data and power, a constraint that has limited earlier bionic eye designs.

The tandem retina combines blue-, green- and red-sensitive layers spanning 300 to 800 nanometers and is fabricated with a 3D laser approach on a hemispherical surface. The motion recognition accuracy of 98.6% was measured under the study's test conditions, and the team has not reported performance outside those settings.

The players

The Hong Kong University of Science and Technology (HKUST)

A public research university in Hong Kong known for its engineering and applied science programs, which conducted the bionic eye research.

Nature Materials

A monthly peer-reviewed scientific journal published by Nature Portfolio that covers materials science and engineering research.

The details

The team built the artificial retina with a 3D laser fabrication method that places dense arrays of light-sensitive pixels on a curved, hemispherical surface, which allows a wide field of view. The device is a significant step beyond the team's 2020 first-generation bionic eye, which contained 100 sensing pixels.

Timeline

  1. 2020: The first-generation bionic eye with 100 sensing pixels was published.

  2. November 13, 2026: The study on the high-resolution bionic eye was reported in Nature Materials coverage.

The Tech Race

The new device extends the HKUST team's 2020 first-generation bionic eye, which had 100 sensing pixels, to 367,500 pixels with full-color imaging. The jump in pixel count and the addition of in-sensor motion processing mark a shift toward artificial retinas that can compete with biological vision in resolution and efficiency.

For people who rely on bionic vision, the study points toward devices that could eventually deliver sharper, full-color images while using far less data than current frame-based systems. The results remain at the laboratory stage, so no near-term change in available treatments is implied by the findings.

The takeaway

Processing motion inside the sensor, rather than sending every frame onward, is a practical design principle that could shape future wearable and implantable vision systems. Readers following assistive technology should watch for clinical testing beyond the laboratory results reported so far.

Further reading

For related developments in sensing hardware and machine vision, explore Robotics.

Source note: This article includes information reported by Tech Xplore.