Using AI to Speed Up the Development of Flexible Transparent Films Essential for Mobile Devices

April 23, 2020 by Luke James

Four institutions have worked together to use artificial intelligence (AI) to accelerate the development of flexible transparent films, it has been reported.

The four institutions—Showa Denko (SDK), the National Institute of Advanced Industrial Science and Technology (AIST), the New Energy and Industrial Technology Development Organization (NEDO), and the Research Association of High-Throughput Design and Development for Advanced Functional Materials (ADMAT)—claim that by using AI, they managed to reduce the number of experiments required to produce flexible transparent films which satisfy specific properties to one-twenty-fifth (1/25) or less of those required when using conventional development methods. 


Developing Flexible and Transparent Film

The focus of the researchers has been on conducting AI-based searches for polymers that exhibit the properties required for designing flexible films for mobile devices. The first step of the teams’ research was to produce 27 types of film.

Then, the researchers incorporated chemical information into them, including molecular structures and mole ratios into explanatory variables using a special method: Extended Connectivity Circular Fingerprints (ECFP4) and chose three objective variables—converted transmissivity, breaking stress, and stretch—which have trade-off relations.

The researchers then made the AI learn the actual values of these variables and produced three types of film based on its recommendations. 


A graphical representation of a flexible film used in mobile devices.

A graphical representation of the flexible transparent film developed by researchers, meant to be used in the development of mobile devices. Image used courtesy of SDK / AIST, NEDO / ADMAT


A Superior Film Through the Use of AI 

It was found that the three types of film recommended by the AI trumped all the types of the film initially made by the researchers, with the three AI films exhibiting physical properties superior to their own.

Thus, the researchers claim, they proved that it is possible to substantially shorten the development period of flexible transparent films by using AI and that it is possible to develop films that exhibit superior properties to films that are made by highly skilled and experienced researchers. 

The researchers hope to refine their technology and develop a system where AI is able to suggest ratios of different materials to produce pre-defined target products with enhanced properties over and above that which could be achieved by human designers.