Communicating with Artificial Intelligence Through Tattoos Inked on the Body

Professor Kim Tae-il's Research Team at Sungkyunkwan University Develops Wearable Device for EEG Measurement Using Tattoo Electrodes
AI Robots That Can Correct Errors and Read Human Cues Expected to Emerge

[Asia Economy Reporter Kim Bong-su] A technology that communicates with artificial intelligence (AI) through tattoos on the body has been developed.

On the 24th, the National Research Foundation of Korea announced that Professor Kim Tae-il's research team at Sungkyunkwan University developed a system that enhances the decision-making performance of AI machines through human brainwaves by creating a wearable device for measuring brainwaves based on tattoo electrodes.

Tattoo electrodes (electronic tattoos) are extremely thin and attach to human skin like a tattoo, monitoring biological signals. Their thinness allows the electrodes to penetrate the contours of the skin, such as fingerprints. In particular, they can overcome the noise problems that are the biggest issue in measuring bioelectric signals.

Currently, pointing out malfunctions or conveying correct operational information to commercialized AI machines relies on passive methods such as hand or voice commands, making continuous performance improvement difficult. Recently, studies have been conducted to provide feedback to AI machines using brainwaves, but most have been limited to controlled laboratory environments due to insufficient performance of brainwave measurement devices and dynamic noise problems caused by user movement.

The research team produced tattoo electrodes with a thickness of about 1 micrometer, which is 1/100th the thickness of a human hair, capable of high-quality brainwave measurement, and integrated them with earbuds (close-fitting wireless earphones) to develop a wireless wearable brainwave measurement device.

The tattoo electrodes can be attached to curved skin surfaces such as the face, and their thinness allows them to adhere without discomfort, like skin, overcoming the chronic dynamic noise problem of existing brainwave measurement devices. By using ultra-small brainwave measurement earbuds that weigh about one-tenth of commercial devices and support wireless communication, the possibility of measuring and utilizing brainwave signals without restrictions even when the user is moving was opened. Additionally, a deep learning model was trained to collect brainwave patterns generated when users observe errors in AI automated machines through the earbuds and extract error-related potentials (specific brainwave patterns occurring upon error detection). As a result, AI machines could confirm error-related potentials in real time and immediately correct errors or enhance machine performance to avoid causing errors.

An AI machine that automatically detects error-related potentials is similar to learning how to read social cues, a human social ability. It can analyze the user's brainwaves by itself to determine positive or negative intentions and learn and decide actions according to the user's desired direction.

Professor Kim said, "Through material technology and engineering efforts, we have taken the performance of wearable brainwave measurement devices to the next level and applied it to the practical field of AI." He added, "We expect the possibility of applying brain-AI automatic control systems to real autonomous vehicles with many variables through follow-up research."

The research results were published online on the 30th of last month in the international journal npj Flexible Electronics.

This content was produced with the assistance of AI translation services.