TUI Group, one of the world’s leading tourism groups, is implementing machine learning in collaboration with Mobi Systems, the US developer of solutions to extend the usability of mobile devices, to optimise scheduling of its destination transfer operations.

Pre-pandemic, TUI sold over 31 million transfers annually, transporting customers between airports, hotels and points of interest in destinations around the world. While transfer scheduling is already automated through TUI’s global transfer operations platform, it will now be enhanced through additional machine learning capabilities meaning “what previously took hours can be done in seconds”.

The application of machine learning to its transfer operations, says TUI, continues its investment in digital platform capabilities and is part of a major push to increase efficiency, enhance services and support sustainable development.

The new transfer platform will integrate real-time data for flights, traffic, weather, vehicle inventory and customer bookings – the key variables that influence transfer schedules.

Mobi’s cloud-based machine learning technology will continuously crunch the data to calculate the most efficient transfer plans in real time, updating schedules instantly and automatically as circumstances change. In addition to substantial time savings, the approach, says TUI, will significantly reduce operation costs and help cut CO2 emissions through greater vehicle optimisation and more efficient transfer routing.

“Integrating new technologies, such as machine learning, helps ensure we deliver the best customer experience through having a faster, more stable and more accurate platform,” said Peter Ulwahn, Chief Digital Officer of TUI Musement, the Tours & Activities division of TUI Group. “The huge advantage of this system is that it can scale to schedule the millions of transfers we manage, while also enabling us to deliver a personalised customer experience.”

Through working with Mobi, the TUI app will also be upgraded to include live vehicle tracking, expected arrival/departure times, and instant notifications of changes to a transfer plan. These features will complement existing functionality, such as advance information on timing and pick-up location of all transfers. The machine learning technology was tested during 2021 with the rollout in key Mediterranean destinations in the coming months.