Driver Scheduling With Rolling Horizon Reoptimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing driver scheduling systems are inefficient due to long computation times, leading to obsolete schedules, especially when using computer-assisted methods on less powerful devices, which fail to adapt to dynamic changes such as driver absences or vehicle unavailability.
Innovation Solution
A system utilizing linear programming optimization with rolling horizon optimization and column generation to rapidly generate and update driver schedules on both mobile devices and cloud infrastructure, reducing processing time and ensuring schedules remain relevant.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional scheduling algorithms are used to optimize driver schedules, then scheduling completeness can be achieved, but computation time becomes excessively long (hours or days)
Solution Approach 1:
The patent segments the scheduling problem into discrete time periods (e.g., 15-minute intervals) and uses dynamic programming to solve each segment sequentially. This breakdown allows the system to achieve comprehensive scheduling coverage without requiring exhaustive computation across the entire scheduling horizon at once, thus reducing total computation time while maintaining scheduling completeness.
Solution Approach 2:
The system implements dynamic programming that adapts to changing conditions in real-time. When events occur (driver absences, vehicle unavailability), the algorithm dynamically recalculates only the affected portions of the schedule rather than reoptimizing the entire schedule from scratch, maintaining scheduling completeness while minimizing computation time through selective recalculation.
2Ease of operation
If optimization is performed on mobile devices with less powerful processors, then scheduling accessibility is improved, but computation speed decreases
Solution Approach 1:
By segmenting the scheduling problem into smaller time-based segments and using efficient dynamic programming transitions, the patent reduces the computational complexity to a level that can be handled by mobile device processors. This segmentation allows the algorithm to run on less powerful devices while still providing comprehensive scheduling optimization.
Solution Approach 2:
The system changes the parameter representation of the scheduling problem, using discrete time intervals and state-based dynamic programming rather than continuous optimization. This parameter transformation reduces computational requirements, enabling the algorithm to execute efficiently on mobile devices with limited processing power while maintaining scheduling quality.
3Manufacturing precision
If schedules are optimized in advance, then scheduling thoroughness is improved, but schedule freshness deteriorates (schedules become obsolete)
Solution Approach 1:
The patent implements a dynamic scheduling system that continuously monitors for changes in driver availability, vehicle status, and other scheduling parameters. When changes are detected, the system automatically recalculates affected schedule segments using dynamic programming, ensuring that schedules remain fresh and reliable while maintaining thorough optimization through systematic recalculation of impacted portions.
Solution Approach 2:
The system incorporates feedback mechanisms that detect schedule disruptions (driver absences, vehicle unavailability) and trigger automatic recalculation. This feedback loop ensures that schedules are updated in response to actual conditions, maintaining both thoroughness through systematic recalculation and freshness through timely updates based on real-time information.
Data Source
AI summary
Systems and methods including one or more processors and one or more non-transitory storage devices storing computing instructions configured to run on the one or more processors and perform: receiving a request to generate one or more schedules for one or more drivers; determining one or more respective day cab schedules for each respective day cab driver of the one or more drivers; assigning one or more permanent drivers of the one or more drivers to one or more permanent tractors; assigning at least one driver of one or more remaining drivers of the one or more drivers to at least one tractor using a first set of rules; generating the one or more schedules for the one or more drivers; and coordinating displaying the one or more schedules for the one or more drivers on an electronic device of a user. Other embodiments are disclosed herein.