Despite their minuscule size, insects are one of the most potent creatures and are even the major idols of robotics and AI. Insects can thrive with minimum energy usage and effective sensing capabilities, all of which are aspects of admiration for autonomous robots.

Recently, Guido de Croon and his team at TU Delft have pioneered a new approach to AI for autonomous robots, drawing inspiration from the intelligence of insects. With the analysis of insects’ neuronal activity when sensing objects, the team created a small robot that uses the same model to navigate a complex area. Ultimately, it was discovered that the main principle of discovery of navigation is to steer towards the area with the least apparent motion. Simply put, when an object appears to be approaching faster relative to others, the insect detects it as closer.

https://www.science.org/cms/asset/d8aefa96-7a1e-4169-8425-1ed7d5f4c0cb/keyimage.gif

The researchers underscored that insights from natural insect intelligence provide a promising foundation for the development of efficient AI for autonomous robots. They anticipate the fostered development of specialized hardware such as a tiny, energy-efficient chip, designed for the ubiquitous use of robots.

Leveraging insights from natural insects opens up possibilities for autonomous agents. With a dedicated sensing capability and optimal navigating algorithm from insects, robots give humans the ability to discover challenging and unknown areas with safety and efficiency. At last, this case has demonstrated that nature is a rich source of inspiration for technological innovation.

인수호 강남포스트 학생기자  webmaster@ignnews.kr

<저작권자 © 강남포스트, 무단 전재 및 재배포 금지>