Virtual Sensor

Virtual sensors leverage existing sensor data and system knowledge to estimate unmeasurable parameters or infer system states, overcoming limitations of physical sensors in terms of cost, placement, or measurement capabilities. Current research heavily utilizes machine learning, particularly graph neural networks (GNNs) and convolutional neural networks (CNNs), to model complex spatio-temporal dependencies within heterogeneous sensor data for accurate estimations across diverse applications. This technology is significantly impacting various fields, enabling improved condition monitoring, predictive maintenance, and enhanced system performance in areas such as industrial automation, autonomous vehicles, and human-computer interaction.