ML Spot Pricing for Air Cargo Logistics Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems lack an efficient and automated solution for spot pricing in air cargo logistics due to the complexity and scale of routing networks, with manual or standard rate sheet-based methods failing to account for the variability in routes and customer behavior, leading to a lack of off-the-shelf tools capable of handling the complexity of global logistics networks.
Innovation Solution
A machine learning-based system that clusters original-destination routes into subgroups based on similarities and customer behavior, determines influencing criteria, and generates price elasticity curves to automate spot pricing, using hardware processors to receive and process data, and visually display results in a graphical user interface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If manual or standard rate sheet-based pricing methods are used, then simplicity of operation is maintained, but the system cannot handle the complexity and scale of global logistics networks with large number of routes and variables
Solution Approach 1:
The patent segments the large-scale logistics network into multiple clusters of routes based on geographical and operational similarities. This segmentation allows the system to manage complexity by dividing the global network into smaller, more manageable units that can be processed independently while still capturing overall network effects.
Solution Approach 2:
The patent introduces a new dimension of analysis by clustering routes not just geographically but also by customer behavior patterns and pricing characteristics. This multi-dimensional approach enables the system to handle complexity by organizing routes in multiple hierarchical levels, transforming the problem from a flat, overwhelming dataset into a structured, multi-layered framework.
2Productivity
If automated machine learning-based pricing system is implemented, then productivity and revenue optimization are improved, but the extent of automation increases system complexity
Solution Approach 1:
The patent implements a self-training machine learning system that automatically learns from historical pricing data and network traffic patterns without requiring manual reconfiguration. The system performs self-service by continuously adapting its pricing models based on observed customer behavior and network conditions, reducing the need for human intervention while maintaining high automation levels.
Solution Approach 2:
The patent incorporates feedback mechanisms where the automated pricing system continuously monitors pricing outcomes, customer responses, and network traffic. This feedback is fed back into the machine learning models to refine pricing strategies, creating a closed-loop system that improves productivity through iterative optimization while managing automation complexity through data-driven adaptation.
3Measurement precision
If clustering of routes into subgroups is performed, then measurement precision of customer behavior patterns is improved, but the difficulty of detecting and measuring increases
Solution Approach 1:
The patent performs preliminary clustering of routes into broader groups based on readily available geographical and operational data before applying more sophisticated customer behavior analysis. This preliminary action reduces the complexity of subsequent measurements by organizing data into meaningful groups first, making the detection and measurement of customer behavior patterns more manageable and precise.
Solution Approach 2:
The patent implements a nested clustering structure where routes are first grouped into clusters, then clusters are further grouped into subgroups based on customer behavior patterns. This nested doll approach allows the system to achieve high measurement precision by applying different levels of analysis at appropriate hierarchical levels, reducing overall data processing complexity through structured organization.
Data Source
AI summary
A machine learning algorithm is trained to learn to cluster a plurality of original-destination routes in a network for transporting cargo into a plurality of clusters based on similarities of the original-destination routes, and to learn to cluster the plurality of clusters into a plurality of subgroups based on customer behavior. Influencing criteria associated with each of the subgroups may be determined and based on the influencing criteria, a price elasticity curve for each of the subgroups may be generated. Based on the price elasticity curve and current network traffic, cargo transportation price associated with each of the subgroups may be determined.