An international team led by Australian scientists has developed the first self-calibrating photonic chip, which can be “transformed” into a bridge on the data highway, change the current connection between optical chips, improve the speed of data transmission, and is expected to promote artificial intelligence and developments in areas such as self-driving cars. The latest research was published in the journal Nature Photonics.

Photonic circuits are capable of manipulating and directing optical channels through which information is transmitted, as well as providing computational capabilities such as searching for patterns that underlie many applications such as medical diagnostics, autonomous vehicles, Internet security, and more. Fast and reliable reprogramming of the chip can speed up the search, but it is very difficult and expensive to do so, and the latest self-calibrating chips overcome this challenge.

A key challenge of this research is to integrate all optical functions into a device that can be “plugged” into existing infrastructure. The solution proposed by the research team is to calibrate the chip after it is manufactured, that is, to calibrate the chip using an integrated reference path rather than an external device, which provides all the settings and switching functions required to “dial”.

Lead researcher, Professor Arthur Lowery, Monash University, said: “We demonstrate a self-calibrating programmable photonic filter chip, which is important because it enables tunable photonic integrated circuits to be widely used in many fields such as optical communication systems that swap signals according to color, extremely fast correlators, scientific instruments for chemical or biological analysis and even astronomy.”

Lowery said that in 2020, the school developed a new type of optical micro-communication chip, which built multiple channels of the data highway and achieved the fastest internet speed at the time. And new self-calibrating chips could be on-ramps, off-ramps and bridges for these data highways, linking those channels and allowing more data to move faster.

The researchers say this latest breakthrough promises to accelerate the development of artificial intelligence for real-world applications such as safer driverless cars that can interpret their surroundings in a timely manner, artificial intelligence that can diagnose conditions more quickly, and smaller photonic network switches, etc.