Photo/Illutration A robot programmed to convert induced pluripotent stem (iPS) cells into retinal pigment epithelium cells is shown in Kobe on June 27. (Fumi Yada)

Japanese researchers combined artificial intelligence (AI) and robotics technology to develop a fully automated system to find the best culturing method for turning induced pluripotent stem (iPS) cells into human tissue.

A team comprised mainly of individuals from the Riken research institute said the android technology will not only produce cells for use in regenerative medicine efficiently but also accelerate bioscience research, saving them loads of time.

“The roles of humans and robots will be increasingly separated with our invention,” said Genki Kanda, a senior researcher at Riken. “Robots will deal with tasks they are set, enabling people to take on more challenging things.”

iPS cells are used to develop all kinds of human tissue, but the process until now has required a hands-on approach in targeting tissue and organs for treatment and other purposes.

As miniscule differences in engineering can massively affect the quality of treated cells, it is difficult for skilled researchers’ techniques to be shared to any great degree and still ensure high quality is maintained.

To overcome the challenge, the team used AI and a humanoid for biological experiments to identify the ideal conditions for creating retinal pigment epithelium cells, which supply nutrients to light-sensing visual cells, from iPS cells.

No human intervention was permitted. The robotic system was programmed to trace the conventional manual cell growing method by scientists.

With a total of 144 scenarios considered, AI’s optimization algorithm assessed culturing results and ended up deciding the best cultivation technique.

According to the team, finished cells boasted as high a quality as those made by veteran engineers.