Biohybrid organoid-robot sensing for olfaction intelligence

Owing to the remarkable advancements in artificial intelligence (AI), the sensory modalities in artificial systems that characterize human embodiment have received significant attention. However, olfaction remains largely absent in artificial systems, primarily due to several technological challenges. In this study, we aim to advance biomimetic olfactory processing by developing a biohybrid organoid-robot (BOR) system. This system integrates an olfactory organoid-based bioelectronic nose, machine learning (ML) decoders, and an odor-triggered robotic platform. By harnessing the sensitivity and specificity inherent in biological olfactory systems, organoid-based bioelectronic noses present a distinct advantage over traditional electronic noses, facilitating the detection of a wide spectrum of odors at low concentrations with rapid response times. Real-time ML-powered decoding of sensing signals triggers predefined actions in the robotic system, thereby establishing a perception-interpretation-actuation loop that enables the BOR system to detect environmental olfactory cues and execute corresponding physical actions. The research presented herein advances the field towards realizing the sense of smell in systems with truly embodied intelligence.