AI-Based Olfactory Sensing for Robotic Platform

This paper presents an AI-based olfactory sensing system designed for robotic platforms to enhance military, defence and security applications to detect threats of terrorist movements or existence of hazardous substances. Although traditional robotic sensory systems such as vision, touch, and hearing have significantly improved situational awareness, their effectiveness is limited in complex, obscured, and nonline-ofsight environments. The proposed system incorporates advanced olfactory sensors and machine learning (ML) models for the detection and classification of hazardous substances such as combustion fuels, smoke, alcohol, food items and toxic gases which indicate the presence of terrorist activities or hazardous substances. A prototype has been developed using edge device ESP32 and grove multichannel gas sensor to demonstrate the feasibility of this novel approach. Achieved over 85% prediction accuracy with target samples. Machine learning models such as Artificial Neural Networks (ANN) and Multinomial Naive Bayes (MNB) were employed to analyze sensor data, ensuring high precision and real-time threat detection. The system's ability to integrate seamlessly with other sensory modalities of existing robotic platform, such as visual and acoustic sensors, further enhances its potential to identify terrorist activities and improved real-time threat response capabilities and situational awareness.