Off-Block Time (OBT) accuracy has significantly increased, by using the Veovo prediction engine to sift through big data from diverse sources and identify patterns.
Veovo, the technology company that is powering up airport performance, has partnered with Auckland Airport to predict departure readiness and potential delays accurately. The new solution by Veovo gives advanced visibility of departure disruptions before they happen – providing the airport, ground handlers and airlines with the crucial time needed to act.
Airport performance relies on accurate and timely off-block time estimates for when an aircraft will leave the stand. However, this critical Target Off Block Time (TOBT) milestone is notoriously difficult to estimate and is often missed or late during busy periods. Using big data, historical trends and a machine learning engine from Veovo, Auckland Airport has increased accuracy from 65% to over 80%, more than ten minutes out from aircraft pushback. The next phase is to include more apron data to move off-block time prediction accuracy higher still.
Machine learning excels at finding insights and patterns in large datasets that humans find hard to spot. With better off-block time predictions, we know which flights are at risk of delay and we can take action to reduce the impact – ultimately delivering a better service to our passengers.’
The Veovo ‘airport machine learning at scale’ approach is built on lessons from multiple airport client engagements. Vast amounts of data from various sources are used to train the prediction model including origin, destination, carrier, aircraft type, gate, passenger load, boarding times and airfield condition.
Hayden Davies, GM APAC at Veovo said, “We have been running the machine learning prediction in parallel with traditional Target off-block time (TOBT) estimates for several months, and the results were consistently and significantly better – giving Auckland Airport and its partners’ confidence ahead of the move to live production. We are excited to be able to help the airport significantly boost its off-block time accuracy, which in turn improves departure sequencing and on-time performance, as well as reducing apron congestion and aircraft fuel burn.
The Veovo prediction engine is available as a cloud service that can operate in conjunction with any Airport Operations System and can be used to improve wider operational decision-making processes such as stand and gate allocations.
