daisy; Augmented analysis of aviation performance

Home | Portfolio | DAISY: Augmented analysis of aviation performance

Industries

Solutions

About the client

EUROCONTROL, the European Organization for the Safety of Air Navigation, provides expertise support and operational excellence to the Air Traffic Management community in Europe. The Agency, in its Network Manager role, is very successful at curating data and trends provided by European Air Service Navigation Providers that facilitate consolidated analysis of aviation network performance.

EUROCONTROL continually augments its extensive data lake to provide insights and detect patterns to support faster solutions for identified issues.

Challenge

Controller-Pilot Datalink Communications (CPDLC) is a digital messaging system delivering requests, instructions and clearances between Air Traffic Control centers and flight crew over a radio communications infrastructure. The high variety of interconnected networks, airborne equipage, or Air Traffic Controller utilization of CPDLC makes its deployment a complex endeavor requiring close collaboration supported by continuous performance analysis and improvement. As a consequence, the sheer volume of data to be processed by analysts makes it challenging to derive insights and trends.

EUROCONTROL requested Skymantics to develop a tool set that accelerates and improves the analysis of events in the air/ground datalink communication network. Analysts need to identify symptoms in the network protocol that point at root causes or user behavior so they can issue recommendations for fixes or improvements of the system. Learn more about our work in Air Traffic Management (ATM).

Solution

In response to this requirement, Skymantics delivered the Datalink Artificial Intelligence System Analysis (DAISY) tool. DAISY digs into existing datalink log datasets and applies Machine Learning algorithms to:

The insights provided by DAISY allow datalink analysts in EUROCONTROL to quickly identify repetitive patterns and affected users, which in turns streamlines the process of coordination and solution of identified issues.

The Machine Learning difference

DAISY removed the manual work for analysts to construct dialogues and investigate events in message sequences. There was much trial and error involved until specific events were found in the datasets.

Instead, now analysts are presented with automatically ordered dialogues and categorized patterns of occurrences. Then, analysts can directly select those that fall within the area of interest for visual investigation.

Skymantics implemented a combination of Artificial Intelligence algorithms to support the DAISY processing intelligence. More specifically, supervised and unsupervised Machine Learning were used to automate the construction, scoring and categorization of dialogues into groups of event patterns. Learn more about our work in Artificial Intelligence.

Machine Learning proved critical not to replace the analyst, but to remove the manual classification tasks and to provide insights that narrow down the investigations. In a system based on a coded language or protocol such as CPDLC, the generation of such insights is fast and reliable.

Other success stories

OGC Disaster and Climate Resilience Pilot

OGC Disaster and Climate Resilience Pilot Client The Open Geospatial...