When artificial intelligence systems fail in the financial or engineering sectors, technicians usually have time to make some patches and restart their algorithms. Everything is fine after a few hours at most. AI technology that works in the aviation industry isn't so lucky. Due to the possibility for real catastrophes, software has to be spot on 24 hours a day.
This is quickly redefining what the phrase mission critical means.
Fortunately, a series of new developments are helping to drastically reduce miscalculations and therefore lead to safer air travel on relatively remote routes. One such route that computer scientists have paid special attention to is in Northern Spain.
The Galicia Case Study
While you wouldn't expect a major Christian pilgrimage site to have anything to do with R&D, computer experts have been paying close attention to travelers that attend the Camino de Santiago or Way of St. James in Galicia, Spain. This journey attracts nearly 300,000 people annually, which puts quite a bit of stress on regional airlines. As a result, it's difficult to provide accurate flight information over the Internet.
IT infrastructure in Galicia is undergoing a number of major upgrades. This has put further strains on the existing manual flight data transmission systems. Computer scientists have found that by applying a self-learning AI alongside human operators they've been able to slowly automate these distribution networks without causing any major issues. Eventually, the need for human intervention will decrease until the algorithms are capable of maintaining flight data autonomously.
While this kind of system might work to handle the needs of some small regional airlines, coders have had to resort to more esoteric methods to prevent problems when applying these techniques on a global scale.
Focusing on Personalization Rather than Distribution
Many people would say that it's foolish to trust an intelligent agent to handle flight data on a worldwide scale without some sort of oversight from skilled technicians. Computer scientists have therefore decided to focus their efforts on allowing AI routines to handle flight personalization systems, which in turn can free up overworked personnel to focus instead on these more mission critical tasks.
Air Canada, for example, has incorporated the Ancillary Sales App from IBM into their regular workflows. This has allowed them to provide new services based on collected metrics. While it might not seem like food and beverage service is anywhere near the level of importance as flight data, the fact that an AI handles it means that human staffers don't have to dedicate nearly as many resources to these tasks. As a result, customers get a better level of service while human engineers are able to instead focus on safety as opposed to other facets of the business.
In fact, offshoots of customer data collection software have indeed led the way for the development of safer skies.
Computer Experts Turn to Baggage Screening
Few people would have ever thought that the terms blockchain and baggage screening would have ever been put next to each other in the same headlines. However, AI based around cryptographic ledgers is proving to be surprisingly efficient at finding whether baggage contains illegal items or other things that could cause serious safety and security problems while airborne.
Ledgers can store baggage information so efficiently because they're designed to sequentially store a number of records. Each time a piece of luggage gets checked in, a new record can be made in an existing chain of similar records. Over time, this might even lead to reduced technical requirements on airports since so much of the heavy lifting is accomplished by remote computers.
Aviation Applications Done Right
Some individuals have suggested that there's a possibility AI technology could disrupt travel if applied incorrectly. However, data from the study in Spain as well as those in Canada and in baggage terminals around the world are helping to show that it can alleviate many problems when used in an appropriate situation.