A real‐world data collection framework for a fused dataset creation for joint human and remotely operated vehicle monitoring and anomalous command detection

Unmanned Remotely Operated Vehicles (ROVs) are widely used in safety-critical missions, spanning across monitoring, surveillance, and search rescue missions. In most cases, the human operator controls ROV under stressful conditions harsh environments. The physical mental states of vital as they may cause involuntary movements, issuing unintentional commands due to fatigue or stress. Consequently, a joint monitoring mechanism that will assess operator's can be beneficial preventing potentially catastrophic events, such accidents. To this end, study, we propose data collection framework for creating fused dataset, consisting obtained from both ROV, integrate stress induction stimulate affect operators during We optimise which includes physiological signals inertial measurement unit (IMU) sensor using fusion approaches feature extraction. evaluate dataset through statistical analysis, illustrating difference between characterising normal commands. Finally, optimised original datasets over variety classifiers, demonstrating an accuracy 94% 95%, respectively.