Insects' internal signals read by AI: Cyborg control system proposed with a "harmonious" rather than "commanding" approach

An international joint research group including Doctoral Student Chowdhury Mohammad Masum Refat and Professor Keisuke Morishima of the Department of Mechanical Engineering at the University of Osaka and Associate Professor Mochammad Ariyanto of Diponegoro University in Indonesia, has proposed a new concept called ISC (Insect Synergy Circuit). In this concept, insects and AI cooperate by having the AI read internal biological signals from insects and utilize them for control. The group successfully validated this concept through demonstration experiments.

The findings were published in the ROBOMECH Journal. Research on insect cyborgs has been conducted globally for more than 20 years. In conventional insect cyborg research, control was managed by using only externally observable behaviors, such as "walking" or "stopping," as information sources.

With this approach, however, it was difficult to reflect the physiological responses that insects exhibit toward their surrounding environments into the control mechanism. To verify the feasibility of the new ISC concept, the research group developed a detachable backpack that enables the simultaneous measurement of "heart rate," "neural signals," and "body movement," along with low-load stimulus inputs, without requiring high-burden surgery. Insect cyborgs equipped with this backpack were placed in five different environmental conditions: natural state, UV exposure, chemical exposure, high temperature, and the presence of food.

After training an AI on the data, the AI was able to identify the current environment from the internal biological information with 93% accuracy. Furthermore, it maintained operational classification accuracy even under conditions closer to real-world environments. In addition, navigation experiments were conducted using an apparatus featuring multiple environments.

The non-cyborg group remained in the room containing food and failed to reach the exit. In contrast, the cyborg group utilized closed-loop control, where stimuli were applied only when the AI determined that the insect was calm, and multiple individuals successfully escaped. In the short term, ISC is expected to find applications in areas where small robots struggle to operate, such as biological sensors for searching for survivors inside collapsed buildings, early detection of harmful insects, and environmental monitoring.

In the long term, it is suggested that AI insects could become natural mediators connecting animals, plants, and humans. By having AI understand the entire ecosystem through insects, this technology could support mutual understanding and communication between species, contributing to the construction of an ecosystem where diverse organisms coexist. Furthermore, the framework of ISC is applicable not only to insects but also to other living organisms and various sensor systems, holding promise for development into next-generation cyborg technology where living organisms and AI cooperate harmoniously.

Morishima stated: "Insects are living creatures, and their responses change from individual to individual and from moment to moment. Conventional bio-cyborg research was a one-way street of 'commanding' the animal. This study challenges the first step of 'state-harmonizing bio-hybrid control,' where an artificial system interacts only after first reading the insect's condition.

Using a wearable device, the AI reads the insect's own heart rate, nerves, and movements, and intervenes only when appropriate. The biggest selling point is the shift from 'manipulating' to 'listening.' This is not a conclusion but a beginning, serving as a foundation for future interspecies dialogue."

Journal Information Publication: ROBOMECH Journal Title: Perception-driven control strategy for bio-intelligent cyborg insect DOI: 10.1186/s40648-026-00344-7 IT Biology This article has been translated by JST with permission from The Science News Ltd. (https://sci-news.

co.jp/). Unauthorized reproduction of the article and photographs is prohibited.

This article has been translated by JST with permission from The Science News Ltd. (https://sci-news.co.

jp/). Unauthorized reproduction of the article and photographs is prohibited.