SATS taps AI-driven task-based planning for airport operations

Early deployments show estimated productivity and planning gains of 10%–15%.

SATS Ltd is shifting from flight-based to task-based planning as it expands the use of artificial intelligence (AI) to improve scheduling and coordination across its airport operations.

The approach will allow more detailed planning of individual ground-handling tasks, SATS Chief Digital Officer Sandeep Sakharkar said at the Singapore Business Review Technology Enterprise Summit on 12 November.

In a fireside chat, he added that the group is seeing estimated benefits of 10% to 15% from some of its early AI deployments, primarily in productivity and planning.

“In the past, a person would have had to gather inputs from multiple touchpoints across the process,” Sakharkar said. “Now, AI can take on some of that coordination in a much more agile manner.” 

He noted that approximately 960 flights arrive and depart daily in Singapore, with disruptions such as adverse weather, flight delays, gate changes, and equipment issues requiring operators to make decisions within tight timeframes.

"We don't start with technology; we start with the problems," Sakharkar added.

Beyond task-based planning, SATS is applying AI to improve cargo handling, particularly warehouse space utilisation and shipment movement across its network.

Sakharkar said the applications support more effective planning of cargo intake and dispatch, whilst ground-handling applications focus on forecasting, scheduling, and anticipating operational disruptions.

"On the ground handling side, it's about improving forecasting, planning, and scheduling. It's about anticipating problems before they occur and understanding how they are interconnected,” he said.

SATS is currently conducting two late-stage AI pilots, one at London's Heathrow Airport and another in Singapore.

The group is also looking to expand AI beyond individual applications towards coordinating information and decisions across its global operations.

However, he identified data quality, process standardisation, governance, and workforce change management as key requirements for scaling such applications. “If [AI] is not accurate. It's not going to work effectively." 

Sakharkar also cautioned against moving too quickly towards autonomous operations, particularly in aviation, where safety and reliability remain critical.

"Autonomy, especially in the agentic AI world, has to be earned," he said.

Beyond airport operations, SATS is exploring the solution in corporate functions, including finance and human resources.

The group has introduced initiatives allowing employees to develop their own AI applications and agents under established policies and safeguards.

One application involves conversational AI tools that help employees retrieve information from thousands of internal documents, including standard operating procedures and work instructions.

Another project includes an internal digital assistant that draws on the group's operational knowledge, historical information, and experience to provide more organisation-specific support.

The applications remain at an early stage but are gaining traction amongst employees, Sakharkar added.