## ## ##

 

CASE STUDY

The AI-Powered Airport

Working with ADB SafeGate to Develop New AI-Powered Products and Services

The Client

ADB SafeGate (ADB) is the market-leading supplier of airport management systems providing innovative, fully integrated solutions to support airport development and optimize operational procedures​. In the course of business, ADB had amassed a great deal of operational airport data, for example: gate allocations, display screens and airport operating databases.

The business had been thinking deeply about the future of the its market offering and had developed a compelling vision for the next generation of products and services​. Perhaps the key date in industry calendar is the Passenger Terminal Expo. ADB were keen to develop some of their ideas and test them with customers at the 2022 trade fair.

THE CHALLENGE

ADB SafeGate had identified a key pain-point for their airline customers: so-called Turnaround Time. When passenger aircraft land, a strict schedule is pre-planned with the objective of getting the aircraft airborne again as soon as possible. Time spent at the airport gate costs money.

However, aircraft can be delayed at the gates by a number of factors - many of which, ADB realised, may be predictable in advance using real-time event and operational data that it already holds. Without such advance knowledge of delays, the norm among carriers is simply to overbook gate time - a clear cost inefficiency.

ADB asked Unai whether they could develop a machine learning model capable of predicting turn-around time in time for the forthcoming trade fair.

What Unai Delivered

  • Predictive Model. Unai took receipt of 3 months worth of data from ADB. The team followed the CRISP-DM methodology to design an approach, engineer the features and develop a model that could predict aircraft turn-around times - with confidence intervals. The model was back-tested on a hold-out set and demonstrated excellent predictive capability.

  • Proof-of-Concept Application. To enable ADB SafeGate to demonstrate the new concept at the trade expo, Unai's engineers wrapped the model up into a cloud-deployable application with a web-application to serve as a User Interface.

The Result

The proof-of-concept was completed in time for the 2022 Paris Airport Passenger Expo. ADB was able to demo the aircraft turn-around application to existing and potential customers, and received strong, positive feedback.

Unai is now working with ADB SafeGate to plan the next phase of the project: turning the proof-of-concept into a fully-fledged product, the first of a new breed of AI-powered offerings that ADB plans to take to market.

We chose Unai not only to deliver a market-leading AI solution but to embed their predictive algorithm into an operational system and maximise potential impact for our clients. The team carefully scoped and delivered an MVP within a very challenging timeline, meeting our deadline for a major trade show and helping secure significant interest from potential customers.

We are looking forward to working with Unai to deliver the solution though to production, with many further enhancements planned for the future

Adrian Blakeley. Head of Products & Solutions, ADB SafeGate