Friday, October 13, 2006

An Automated Travel Assistant

Every since I first viewed the historic Knowledge Navigator video produced by John Scully CEO of Apple in the late 1980's, I have believed that an intelligent assistant would truly be a killer app for the travel industry. As originally envisioned, Knowledge Navigator was going to be a tablet, the size of an opened magazine, and would have very sophisticated artificial intelligence. The machine would anticipate your needs and act on them. The Apple Newton was the first generation of these types of tools, but as anyone familiar with technology knows, the Newton was a resounding failure. (Though Palm a few years later did make PDAs a reality).


I have been involved with various AI scientists since the mid 1980s while I was still at United Airlines. A friend (the husband of one of my wife's bridesmaids) worked for a company called MAD Intelligent Systems here in Silicon Valley. Steve and I would spend hours talking about how to apply AI to the travel industry. Sadly Steve passed away in the late 1980s dying of leukemia at the young age of 35. Steve inspired me to continue to explore the possibilities of AI in the travel space.

We are now in the future (certainly in 1980s terms) and AI applications have been used in a variety every day activities. Examples include control, planning and scheduling, the ability to answer diagnostic and consumer questions, handwriting, speech, and facial recognition.

Over the years I have worked with a variety of clients who have tried to use optimization technology, constraint engines and other pieces of AI software to improve the travel process. Disruptive technology such as the Orbitz fare matrix was built by the AI scientists at ITA Software. My partners at Fetch Technologies represent some of the brightest minds in AI today. Once we are funded I plan to use AI techniques to fulfill the dream first described in John Scully's video. The creation of a master itinerary will allow us to run background queries to search for better fares, rates, and travel research items based on consumer preferences. In the early 1990s TMCs touted their mid-office software (such as TRX's CoRRex) as a quality control tool and as a way to search for better fares and seats against the GDS inventory. We are now in a much more fragmented, Web-based environment and the ability to use AI "agents" to perform similar tasks against Web information is not only possible, but can truly change the travel shopping experience.