At any given time, Lufthansa CityLine turnaround coordinators have their eyes fixed on monitors showing more than a half-dozen video feeds of airplanes parked at airport gates. It is the coordinators’ responsibility to ensure that the planes are unloaded, refueled, cleaned, restocked, and reloaded so that every passenger arrives safely, on schedule, and with their luggage.
Minutes lost in the turnaround procedure can quickly add up, costing airlines millions of dollars every year. Many in the industry point out that airplanes only make money while they are in the air.
“Think of a pitstop in a car race, and this is very much what happens in an aircraft turnaround,” said Philipp Grindemann, Lufthansa CityLine’s head of business development and project management. “All processes must be completed on time, quickly, and efficiently.”
Lufthansa CityLine is a subsidiary of Lufthansa, one of the world’s largest airline corporations with a global network. Lufthansa has hubs in both Frankfurt and Munich, Germany. Lufthansa CityLine operates more than 300 flights each day between these hubs and destinations throughout Europe. Arrivals and departures on time are critical for client satisfaction and Lufthansa’s bottom line.
Outside of weather, delays are caused by errors in the closely planned turnaround procedure. Lufthansa CityLine, like most industry participants, relies on manual timestamps to determine when each phase of the turnaround process begins and ends, and uses that data to get insights about where to make improvements for faster, leaner turnarounds.
The airline collaborated in a pilot phase with zeroG, a Lufthansa Group consulting group formed by Lufthansa Systems to speed the practical impact of artificial intelligence in operational and commercial operations at airlines throughout the world. One example is using AI to improve turnaround management.
ZeroG’s Deep Turnaround solution makes use of Azure Video Analyzer, a new Microsoft service that combines Live Video Analytics and Azure Video Indexer capabilities. It creates automatic timestamps from video feeds for Lufthansa and sends alarms when the turnaround strayed from the script.
“With the openness of Deep Turnaround, the airline can direct the process and have significantly leaner processes than before,” stated Manuel van Esch, lead consultant for zeroG.
Deep Turnaround, for example, notifies turnaround coordinators and other ground operations staff when a fuel delivery arrives later than expected. The alarm initiates a search for a solution that will avoid a delay, such as delivering a second fuel truck to the plane.
Azure Video Analyzer is one of several Azure Applied AI Services introduced by Microsoft on Tuesday during Build, the company’s annual developer conference. Azure Video Analyzer, Azure Metrics Advisor, Azure Bot Service, Azure Cognitive Search, Azure Form Recognizer, and Azure Immersive Reader all help to accelerate the creation of scenario-specific AI solutions.
The AI models at the core of Azure AI products and services serve as the foundation for Azure Applied AI Services. Azure Cognitive Services, for example, provide configurable AI models and tools for developing AI solutions that help customers in extracting meaning from text, integrating speech into apps and services, identifying and analyzing content inside photos and videos, and making decisions.
Customers can also customize these services and expand them with their own unique Azure Machine Learning models to match their specific business needs.
Customers usually tell Microsoft that, while they understand the potential of AI, developing solutions is more difficult than expected, according to Eric Boyd, corporate vice president of Microsoft Azure AI in Redmond, Washington.
“With Azure Applied AI Services, the idea is to provide a bit more packaging and structure to accelerate the creation of AI solutions for typical business operations,” he explained.
For example, the Azure Video Analyzer service combines Computer Vision from Azure Cognitive Services and an automatic captioning model, as well as capabilities for integrating existing closed-circuit video feeds and video management systems, making it easier for businesses to build video analytics solutions.
The Azure Applied AI Services category was designed by Microsoft to address typical business scenarios that Boyd’s Azure AI team saw the users consistently build from scratch. Azure Form Recognizer, for example, is based on optical character recognition, a computer vision technique that detects text and is used in a variety of commercial solutions ranging from reading receipts to extracting data from intake forms.
“There was so much more they needed to do to put it in the application that they wanted,” Boyd explained. “It wasn’t just about obtaining the text; it was about comprehending the structure of the document and saying, ‘I have this form that someone has filled out, and I want the information on it in my database,'” explains one.
Azure Form Recognizer extends the underlying optical character recognition technology by providing a framework for comprehending the entire document structure, extracting essential information, and populating a database.
Many Azure Applied AI Services are based on AI tools that were originally created for internal products and services, such as Azure Metrics Advisor. The program evolved from work done by developers for Microsoft’s Bing search engine to detect aberrations from regular operations, such as increases in inquiries from one country or a sudden decline in advertising revenue.
“Search is pretty predictable in terms of how it evolves day to day,” Boyd explained. “By being able to recognize those anomalies, we could more rapidly jump on the problems and rectify them.” “That anomaly detector service has been rolled out to a few places, including Power BI, but it’s a developer’s interface that requires you to link a lot of stuff together.”
Microsoft made the technology public via Anomaly Detector, one of the Azure Cognitive Services. Microsoft built Azure Applied AI Services on the technology that powers Anomaly Detector and tailored it to common solutions for business customers, making it simple to deploy a solution that monitors metrics and, when something goes wrong, issues an alert and flags where to look to resolve the problem.
In China, Samsung Electronics deployed Azure Metrics Advisor for anomaly detection and root cause analysis of problems that could lead to outages on the cloud-based hardware and software system that enables 24/7 access to audio and video content streamed over the internet for display on the company’s Smart TVs.
According to Jie Zhang, technical leader of Samsung Electronics (China) R&D Centre, who assisted with design and implementation, the AI solution built with Azure Metrics Advisor enables Samsung engineers to detect incidents before they affect customers and promptly fix the issues.
According to Boyd, the backend development of Azure Bot Service followed a similar path as Azure Metrics Advisor. To aid customers in developing intelligent conversational assistants, this service leverages the key speech and language technologies that underlie Azure Cognitive Services such as Language Understanding, QnA Maker, Speech to Text, and Text to Speech.
“We aggregate a lot of Cognitive Services and package them together to make it easier for consumers to use the entire combined product,” Boyd explained.
Several of the services now available under the Azure Applied AI Services category were previously available as standalone Azure Cognitive Services, such as Azure Form Recognizer and Azure Immersive Reader, which allows developers to incorporate techniques into their applications that improve reading and writing for people of all ages and abilities.
Other Azure AI capabilities included Azure Bot Services and Azure Cognitive Search, which allow developers to integrate AI-powered search into their apps and websites.
According to Boyd, the restructure is intended to make it easier for corporate clients to find AI solutions to typical business procedures. The new category is expected to grow in the coming months and years as the Azure AI team works with customers in specific industries or sees customers solve similar problems using a combination of Azure AI capabilities.
“We see a category of ‘How can you bundle, mix, and simplify these things?'” “I’m speaking to people that recognize the great power of what AI can do and are saying, ‘I need to apply that to all parts of my organization,'” Boyd said.
The power of AI to generate and interpret data is the driving force behind zeroG’s use of Azure Video Analyzer for the Deep Turnaround solution piloted by Lufthansa CityLine. According to Xavier Lagardere, the aviation company’s chief data officer, this is just the beginning of a digital transformation with AI across the Lufthansa Group.
“When it comes to making choices, decisions, or simply acting on data, we’re not yet systemically a real-time data-driven type of organization,” he said. “There’s an exciting road ahead to do much more with the massive amounts of data that we collect daily.”
Here at CourseMonster, we know how hard it may be to find the right time and funds for training. We provide effective training programs that enable you to select the training option that best meets the demands of your company.
For more information, please get in touch with one of our course advisers today or contact us at training@coursemonster.com