Guwahati: What if a computer could recognise a face or interpret a traffic scene more like the human brain, without having to move data between separate components for capturing, storing and processing images?

Researchers at the Indian Institute of Technology (IIT) Guwahati, in collaboration with IIT Kharagpur, have developed a light-sensitive device inspired by the human brain that combines these functions while consuming 500 times less energy than conventional MXene-based devices in laboratory tests.

The ultrathin device can detect light, retain information even after the light source is switched off, and strengthen its response when exposed to repeated light signals. The researchers say the technology could find applications in facial recognition, robotic vision, traffic monitoring and other intelligent computing systems.

The innovation addresses a key challenge in conventional electronic vision systems, where cameras capture images, memory units store data and processors analyse them. Moving information between these components consumes energy and can slow down processing.

The human brain, by contrast, can recognise familiar objects rapidly by drawing on information already stored through experience. The IIT researchers have sought to reproduce aspects of this ability in an electronic device.

Explaining the difference, Prof. P.K. Giri, lead author of the research and a professor in the Department of Physics at IIT Guwahati, said a conventional system captures light reflected from an apple, converts it into digital information and sends it to a microprocessor for pixel-by-pixel analysis. The human brain can identify the same object by drawing on its learned understanding of colours, shapes and context.

The team’s device seeks to bring sensing, memory and information processing together on a single platform, reducing the need for separate operations.

At its core are two materials: MXene nanosheets, an ultrathin, electrically conductive two-dimensional material, and an organic semiconductor called P3HT, which generates electrical charges when exposed to light.

When the materials are brought into contact, light-induced charges separate at their interface and remain separated for an extended period. This produces a memory-like effect, allowing the device to retain a response after illumination ends.

The device also exhibits a learning-like behaviour. A brief flash of light generates an electrical response that persists after the light is switched off, while repeated light signals strengthen that response. In experiments, the response to a second light signal was 180% stronger than the first.

“This is similar to our neurons learning through repeated stimulation,” Giri explained, describing how the device combines sensing, memory and information processing.

The researchers tested the technology in facial recognition and traffic-scene segmentation, a process in which different elements of a road scene are identified and separated for analysis. The device achieved 90% accuracy in facial recognition and 85% in traffic-scene segmentation tests.

Developed using a low-cost solution-processing technique, the device could offer a route towards more energy-efficient visual computing. Potential applications include intelligent surveillance, autonomous and robotic systems, and other technologies that need to interpret visual information.

The findings have been published in the journal Carbon. The study was co-authored by researchers from IIT Guwahati and IIT Kharagpur, including Giri, Sanjoy Sur Roy, Sanju Nandi, Debabrata Sahu, Garima Choudhary, M. Meyyappan, Saswata Goswami and Chandra Sekhar Tiwary.

While the results demonstrate the device’s potential under laboratory conditions, further development would be needed to establish its performance and suitability for large-scale, real-world applications.