Iterative No-Show Forecasting for Airline Overbooking Decisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional overbooking strategies in the airline industry rely on broad estimating techniques that result in marginally accurate passenger no-show forecasts and cost data, leading to inefficiencies and increased costs due to empty seats and denied boardings.
Innovation Solution
A forecasting and overbooking management system (MARS) that minimizes costs by accurately predicting no-show rates and denied boarding costs, considering factors like voucher amounts, volunteer percentages, and ripple effects, to determine optimal seat allocations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional broad estimating techniques are used for overbooking, then implementation is simple, but forecasting accuracy of no-show rates and cost data is marginally accurate
Solution Approach 1:
The patent segments the overbooking problem into multiple independent probability calculations for different time periods (e.g., 24-hour, 12-hour, 6-hour windows). Each segment calculates no-show probability separately using iterative methods, allowing the system to achieve high forecasting accuracy through divided calculations rather than a single complex model.
Solution Approach 2:
The system dynamically updates no-show probabilities iteratively as time progresses and new information becomes available. The probability calculations are not static but are continuously refined based on actual show-up patterns and remaining capacity, enabling the system to adapt to changing conditions and maintain high accuracy.
2Productivity
If overbooking level is increased to reduce empty seats, then revenue from filled seats increases, but costs of denied boardings increase
Solution Approach 1:
The patent implements a feedback mechanism where the overbooking system continuously monitors actual no-show patterns, denied boarding occurrences, and cost outcomes. This feedback is fed back into the iterative probability calculations to refine future overbooking decisions, allowing the system to optimize the balance between seat utilization and denied boarding costs based on real performance data.
Solution Approach 2:
The system changes the overbooking parameter (number of additional seats sold) dynamically based on calculated no-show probabilities for different time periods. Rather than using a fixed overbooking percentage, the system adjusts the overbooking level by modifying probability parameters iteratively, enabling precise control over the trade-off between filled seats and denied boarding risks.
3Object-generated harmful factors
If overbooking level is decreased to reduce denied boardings, then denied boarding costs decrease, but revenue from empty seats increases
Solution Approach 1:
The patent performs preliminary iterative probability calculations in advance to determine the optimal overbooking level before the flight departs. By calculating no-show probabilities for different time windows ahead of time and simulating various overbooking scenarios, the system proactively identifies the optimal booking level that minimizes both empty seats and denied boardings, rather than reacting to actual show-up patterns after the fact.
4Measurement precision
If iterative probability calculations are performed throughout the period, then forecasting accuracy improves, but computational time increases
Solution Approach 1:
The patent applies periodic action by performing iterative probability calculations at specific time intervals (e.g., 24 hours before departure, 12 hours before, 6 hours before) rather than continuously. This periodic approach maintains high forecasting accuracy by updating probabilities at critical decision points while avoiding unnecessary computational overhead during intermediate periods when no new decisions are required.
Data Source
AI summary
Systems and methods to generate predicted variances of an operation based on data from one or more connected databases.