Late to gate prediction for an airlinesAviation
Out of all operational expenses, the biggest volatile cost to the airline is flight delay. Every year, airlines face millions of dollars in losses and fines due to flight delays, whether it is from climate reasons, technical issues, or late passengers. With each passing minute, airlines have to incur airport charges, parking fees, and adjustments to the schedules of other flights. One factor not directly controlled by airlines is late passengers, which, despite being beyond their control, significantly impacts operational efficiency and costs to the company.
The client, a prominent airline company in the GCC region, was facing operational challenges primarily due to late passengers, an uncontrolled parameter. The major challenges were:
- Identification of key attributes
Determining various external factors associated with customers directly contributing to delay.
- Identification of airport attributes
Identification of airport areas and processes consuming the majority of time.
- Customer Segmentation
Clustering customers into groupsÂ
- Developed a late to gate prediction model capturing information across dimensions like passenger booking, skywards program, trip details, etc.
- The client wanted to identify a segment with around 30% late to the gate which was achieved.
Delivered successful POC resulting in sale of SAS license to the airlines which was an IBM shop. POC showcased following:
- Reduced Operational Cost
- Improved Operational Efficiency
- Enhanced Customer experience