I keep holding to the vague hope that someone, some day will find a way to not just analyze this mountain of data, but interpret it in Operations Terms and present it in a meaningful way to people actually running the plant. Not a network event, but a plant event. Am I dreaming?