Vehicle Route Optimization via Geographic Partitioning and Dynamic Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle routing problem (VRP) solutions struggle to efficiently service millions of customers in a larger geography within a planning period with scalable hardware and system configurations, failing to address growing customer demands and business requirements while minimizing manual effort and costs.
Innovation Solution
A system and method for vehicle route optimization that includes creating initial routes based on network geometry, applying heuristics, and exchanging orders between nearby routes to optimize delivery schedules while adhering to time windows and capacity constraints, using a combination of hardware and software components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing VRP algorithms are used to service customers, then delivery routes can be planned, but the system cannot efficiently handle millions of customers in large geography with scalable hardware
Solution Approach 1:
The patent divides the large-scale VRP problem into multiple smaller sub-problems by partitioning the geography into zones and processing routes in parallel across multiple computing nodes. This segmentation enables the system to handle millions of customers by breaking down the computationally intensive optimization into manageable chunks that can be distributed and processed concurrently, directly addressing the scalability limitation.
2Productivity
If more vehicles are deployed to service more customers, then customer demand is met, but logistics cost and travel distance increase
Solution Approach 1:
The patent employs advanced optimization algorithms that dynamically adjust routing parameters, delivery time windows, and vehicle capacity allocations to minimize total travel distance while meeting customer demand. By optimizing these parameters rather than simply adding more vehicles, the system achieves higher productivity with reduced logistics cost and energy consumption.
3Productivity
If optimization is performed for NP Hard VRP problems, then route efficiency improves, but computation time becomes unreasonable
Solution Approach 1:
The patent performs preliminary actions by pre-processing customer data, pre-defining geographic zones, and pre-establishing routing templates before the actual optimization run. This preliminary structuring of the problem reduces the computational complexity during the optimization phase, enabling efficient solution of NP Hard VRP problems without excessive computation time.
Solution Approach 2:
The patent implements dynamic optimization approaches that adapt the level of optimization detail based on problem size, time constraints, and computational resources available. For large-scale problems, it uses hierarchical optimization where high-level routing decisions are made quickly and detailed route optimization is performed selectively, balancing route efficiency with computation time.
4Reliability
If delivery time windows are strictly enforced, then customer satisfaction improves, but route flexibility and optimization difficulty increase
Solution Approach 1:
The patent applies different levels of time window constraint strictness to different customers or delivery locations based on their importance, urgency, or historical performance. Critical deliveries have strict time window enforcement, while less critical ones allow more flexibility. This local differentiation maintains customer satisfaction for high-priority deliveries while reducing overall optimization complexity.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
State of the art vehicle route optimization approaches fail to handle growing customer needs. The disclosure herein generally relates to route planning, and, more particularly, to a method and system for vehicle route optimization. The system obtains information on (i) a network of customers, wherein the network of customers comprises nodes corresponding to a plurality of customer locations and a depot, and edges representing connections between the customer locations, (ii) distance between each pair of customer locations among the plurality of customer locations, (iii) one or more orders received from one or more of the customers, (iv) service start time, (v) service end time, and (vi) servicing time, as input data, and then by processing the input data, performs vehicle route optimization at different levels, till an optimized route plan is obtained.