Pressure Sensor Data Modeling with Recurrent Conditional Generative Adversarial Networks

Several critical systems in an autonomous car strongly rely on pressure sensors to track and measure critical specifications, which has winded up to be the crucial aspect in making secure roadways, enhancing the driving experience, and minimizing pollution. Autonomous car driving experience would be completely different without all types of pressure sensors employed in the contemporary car, these sensors aid to handle braking systems to auto electrical windows, exhaustive emissions to the power steering. Monitoring the health of these pressure sensors is done by physically collecting the time series data at different scenarios. This approach is expensive and time-consuming. As a result, autonomous vehicle corporations are focusing on virtual authentication where synthetic data is generated for several scenarios. RCGAN is exploited in this work, for generating time series synthetic sensor data. This algorithm employs RNN, which trains the network model on supplementary information. This change permits RCGAN to learn and create sensible, time-series sensor data.