Despite all efforts in realising the perfect universal base comfort level, there is still a considerable amount of interference with the controls of a automatic HVAC system by passengers. To provide personalised comfort levels to individual passengers a truly learning system based on a neural network can be successfully used. This system in combination with a passenger behaviour management system adapts the factory-set base comfort level to the individual preferences thus reducing the amount of interference and increasing the individual satisfaction with an improved HVAC control system.