On the 21st, Golem, a tech news website, published a blog post reporting that a team at the Tokyo Institute of Science has achieved a technological breakthrough, successfully converting LED light energy into electrical energy for the first time, realizing wireless power supply without batteries or cables.
According to the report, this technology belongs to the field of Optical Wireless Power Transmission (OWPT). Its basic principle is to convert electrical energy into light energy for transmission, and then a photovoltaic receiver converts the light energy back into electrical energy. Unlike previous laser-based solutions, this new technology uses high-power LEDs, providing a more promising path for powering indoor devices.
The core advantages of this technology lie in its high safety and low cost. In indoor environments with a high density of IoT devices, wireless power transmission systems must adhere to strict safety regulations to avoid harm to the eyes and skin.
Traditional laser solutions, due to their high energy density, cannot meet these requirements, while LED-based technology is inherently safer. The research team points out that this characteristic makes it ideal for building sustainable infrastructure for indoor IoT devices and enables simultaneous, uninterrupted power supply to multiple targets using AI image recognition.
To overcome energy loss and performance fluctuations under varying lighting conditions during long-distance LED wireless power transmission, the research team developed a dual-mode adaptive system capable of automatically adapting to both bright and dark indoor environments.
The key to this system lies in an adaptive optics system composed of a tunable liquid lens and an imaging lens. This system automatically adjusts the beam size based on the distance and size of the receiver, ensuring optimal energy transmission efficiency.
For precise beam positioning, the system integrates a depth camera and an adjustable reflector controlled by a stepper motor. The RGB sensor in the depth camera identifies the location of the photovoltaic receiver, while the infrared sensor locates the beam's illumination point.
Furthermore, the researchers attached a retroreflective film to the receiver's edge, reflecting the infrared light from the depth camera. This allows for clear contouring of the receiver even in complete darkness, ensuring stable operation of the system around the clock.
The research team further introduced a convolutional neural network (CNN) based on the SSD algorithm, significantly improving the accuracy of target recognition. In the experiment, the system demonstrated seamless operation in both bright and dark environments, successfully achieving efficient and stable energy transfer over a distance of up to 5 meters. According to the research report, the LED chip used in the system has a radiant flux of 1.53 watts.