NASA scrubbed the space shuttle Discovery’s launch yesterday, with less than two and a half hours to go. Why? One of the fuel level sensors in the big hydrogen tank stubbornly refused to acknowledge that the tank was in fact full.
As if underscoring the importance of sensors, NASA’s spokespeople said they have an “unexplained anomaly” on their hands. In other words, they can’t figure out why the sensor keeps dying.
This large and public endorsement of predictive maintenance (evil grin) should make us all reflect on the importance of sensors to process automation.
Yes, we are all thinking on a higher level today: optimization, integration, interoperability, and so forth. But the heart of any business information system, whether ERP-based or plant floor-based is the collection of information. On the plant floor, sensors and final control elements are the face cards. If your sensors don’t work, your system won’t control, your historian is worthless, and the data you send to the business system is bogus.
Let’s hope NASA figures out their problem so that Discovery can launch soon. And let’s hope that we can find the kind of predictive maintenance tools we need to keep our plants working and our data good.