Describir: Model-Based Prediction of Operation Consequences When Driving a Car to Compensate for a Partially Restricted Visual Field by A-Pillars