AI technology paves the way for wireless power transmission: A major step towards a 'wireless' society.

Wireless power transfer (WPT) has long been present in everyday life, from charging smartphones and biomedical sensors to induction cooktops. However, this technology still has a major limitation: the output voltage is prone to fluctuations when the load changes, leading to reduced charging efficiency and the potential risk of damaging devices.

The reason lies in the fact that most current WPT systems are load-dependent, meaning the efficiency of power transfer is heavily influenced by the device being powered. With phone batteries, resistance changes as the charge level increases, leading to unstable voltage, slowing down the charging process, or even affecting battery lifespan.

A Japanese research team has introduced a completely new approach: using machine learning technology to design a Load Independent (LI) WPT system. This allows the system to maintain stable voltage and high efficiency regardless of changes in the characteristics of the device being charged.

In the context of phone battery charging, this means that energy is transferred continuously and safely, even when the battery is nearly full. This benefit becomes even more important for larger and more complex applications, such as electric vehicle batteries, where the load can vary dramatically throughout the charging cycle.

How does a wireless power transmission system work?

Essentially, WPT is based on the phenomenon of electromagnetic resonance, similar to how radio waves or TV receive signals. The transmitter generates electromagnetic waves at a specific frequency by oscillating energy between an inductor and a capacitor. When the receiver is tuned to the same frequency, the two circuits resonate, amplifying the transmitted energy.

In wireless power transmission systems, this resonance process allows the receiver to absorb and store energy instead of decoding the signal as in broadcasting equipment.

The breakthrough in the research lies in how AI directly participates in the design process. Scientists built a virtual model of the WPT system, then allowed the AI ​​to observe and evaluate its performance through a series of simulations.

Artificial intelligence continuously analyzes factors such as energy loss in the form of heat, voltage stability, and the "cleanliness" of the electrical signal. Through trial and error, AI gradually optimizes the circuit structure, helping the system achieve the highest efficiency with the lowest fluctuations and energy losses.

The results showed that voltage fluctuations were reduced to just 5%, compared to 18% in traditional load-dependent systems. At the same time, power transmission efficiency reached 86.7%, far exceeding the lowest level of approximately 65% ​​in older designs.

A step toward a completely wireless society.

The research was published in June 2025 in the prestigious international scientific journal IEEE Transactions on Circuits and Systems, a leading journal in the field of electrical and electronic engineering. Professor Hiroo Sekiya, the lead author and lecturer at Chiba University (Japan), believes that the significance of the work goes far beyond the story of conventional wireless charging.

“We believe this is a significant step towards a truly wireless society,” emphasized Hiroo Sekiya. “Thanks to its load-independent operation, WPT systems can be designed to be simpler, more compact, and more cost-effective. Our goal is to bring this technology into widespread application within the next 5-10 years.”

Not only does the research open up a new future for wireless power transmission, but it also shows that AI is changing the way humans design electrical circuits, gradually moving towards an era of automated power electronics design, where AI plays a central role in technological innovation.