Vehicle Routing and Capacity Utilization for Incompatible Commodities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing logistics systems face challenges in simultaneously optimizing vehicle routes and capacity utilization, especially when dealing with heterogeneous commodities and incompatible commodities, leading to increased computational complexity and inefficiency in quantum hybrid solvers.
Innovation Solution
A method involving splitting pickup locations into nodes based on commodity type and capacity, clustering nodes geographically, and assigning vehicles iteratively to each cluster, followed by an optimization model solved using a quantum hybrid solver to minimize an object value, thereby optimizing routes and capacity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If quantum hybrid solvers are used to solve logistics optimization problems, then solution quality improves, but computational time increases due to increased number of variables and constraint equations
Solution Approach 1:
The patent divides the logistics optimization problem into two separate sub-problems: vehicle routing problem (VRP) and vehicle loading problem (VLP). By segmenting the original complex optimization model into smaller independent sub-models, the number of variables and constraint equations in each sub-problem is significantly reduced, allowing quantum hybrid solvers to solve them more efficiently within reasonable computational time while maintaining solution quality.
2Productivity
If systems handle both vehicle routing and vehicle loading optimization, then logistics efficiency improves, but the number of variables and constraints increases leading to more computational steps
Solution Approach 1:
The patent separates the integrated logistics optimization into distinct VRP and VLP modules. The VRP module handles route planning with variables related to vehicle positions and routes, while the VLP module handles commodity loading with variables related to cargo placement. This segmentation reduces the joint complexity that would arise from combining all routing and loading decisions in a single model.
Solution Approach 2:
The patent performs preliminary route planning in the VRP stage before optimizing loading in the VLP stage. By pre-determining vehicle routes and assignments in the first stage, the second stage only needs to optimize loading configurations for fixed routes, significantly reducing the search space and computational complexity compared to simultaneously optimizing both routing and loading.
3Reliability
If heterogeneous commodities are handled with separation of incompatible commodities, then logistics safety improves, but quantum hybrid solver complexity increases
Solution Approach 1:
The patent applies different handling rules to different commodities based on their compatibility characteristics. The VLP model incorporates commodity-specific constraints that identify incompatible pairs and prevent them from being loaded into the same vehicle. This localized differentiation approach ensures safety requirements are met for each commodity type without uniformly increasing complexity across the entire optimization model.
Data Source
AI summary
Logistics planning with optimal route planning and optimal vehicle capacity utilization, while addressing incompatible commodities is NP-hard problem. A method and system for optimal route planning and vehicle capacity utilization with management of incompatible commodities in logistics is disclosed. For high volume of commodities exceeding vehicle capacity at a given pickup location and need of separating incompatible commodities present at same pickup location across vehicles, the locations are split into nodes. Further, the multiple nodes are clustered using a customized clustering technique and a first level vehicle assignment is performed per cluster. Furthermore, with the created nodes and assigned vehicles per cluster an optimization model defined by an object value (O) is created for each cluster. It is solved using a quantum hybrid solver that minimizes the object value under defined constraints. The solution obtained provides an optimal route with optimal capacity utilization for each vehicle for each cluster.