I’ll Take Troubleshooting for $500 Alex
I’m a fan of Jeopardy, the TV trivia show hosted by Alex Trebek. Contestants in the show compete for monetary prizes based on how they respond to the categories in increasing dollar amounts and how they wager when a Double Jeopardy clue is found. A big part of winning is also how they wager and respond to the Final Jeopardy question. It’s a fun way to learn about subjects I would have never considered exploring.
In 2011, Watson, a computer developed by IBM that uses artificial intelligence (AI) to answer questions posed to it in natural language, competed against former Jeapordy winners Brad Rutter and Ken Jennings on the show. Watson won the competition! You can click here to learn more about Watson on Wikipedia.
Artificial intelligence is all around us. Several commercial applications are in use. IPhone users talk to Siri, Android users pose questions to Google and others pose questions and commands to Amazon’s Alexa.
Artificial Intelligence is now infiltrating hangars around the world. On December 12, Aircraft Technical Publishers (ATP) announced the acquisition of CaseBank Technologies. CaseBank computer-based troubleshooting system is called SpotLight. SpotLight uses what CaseBank calls “hybrid reasoning” that pulls information from technical data and past discrepancies and solutions to quickly troubleshoot faults based on trends across aircraft fleets. As more aircraft mechanics use SpotLight, it’s knowledge and troubleshooting capabilities increase dramatically. In addition, SpotLight pulls information from all users, so even a small shop can gain a significant troubleshooting benefit based on the experiences of other mechanics in the field. It’s a way to seamlessly share tribal knowledge with all users.
The first step of traditional troubleshooting is to understand how the system works in a normal mode so that we can then figure out what the critical paths of influence are. We then test different parts of the system to determine the component at fault. SpotLight takes care of those first steps. Its knowledge is based on OEM engineering information and user discrepancies and solutions. When a discrepancy is entered, SpotLight starts to ask the mechanic a series of questions to pinpoint the solution.
I was able to get a demo of SpotLight and must say I was impressed with its troubleshooting abilities. I was also impressed with the ability of SpotLight to help uncover other maintenance issues during routine/heavy maintenance. In the example we went through, a mechanic was assigned to inspect the aircraft’s VHF antenna. Upon inspecting the antenna, she finds deteriorated sealant. She then enters this information into the system. SpotLight then searches its information database to see if there are any past discrepancies that could be related to this defect. When the mechanic clicks on the repair procedure, SpotLight notes that based on past discrepancies in the field, water leaking from bad sealant on VHF antenna has led to problems with flight attendant call buttons. Leaking water gets to the terminal block behind the call button and bridges the isolation gap, making it seem as if a flight attendant is holding down the call button. It then queries the maintenance database of the aircraft being worked on and finds recurring problems in the past regarding the flight attendant call button. The maintenance department now knows that this problem has likely been fixed by correcting the sealant issue on the antenna.
The question begs to be asked – how will this advancing technology affect the actual troubleshooting skills of mechanics?
Well, how has GPS navigation affected our ability to read and understand maps? I would argue that those who already knew how to read maps can now get from Point A to Point B easier, while those who never learned to read a map can just as easily get from Point A to Point B.
The problem comes when there is no GPS available. Then what?
The same goes for troubleshooting!
Thanks for reading, and we appreciate your comments.