Dynamic Disk Travel Time Prediction for Paratransit Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current travel time prediction systems for paratransit demand-response models face inefficiencies and inaccuracies due to obstacles and varying speed zones, leading to increased processing time and potential delays in scheduling, which can result in costly financial penalties and customer complaints.
Innovation Solution
The system represents addresses as disks dynamically sized to encompass a geographic area, allowing for accurate and efficient travel time predictions by using a disk-to-disk lookup table, which reduces the number of calculations and improves memory usage, thereby enhancing processing speed and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional grid-based methods are used to represent geographic areas for travel time prediction, then coverage of geographic area is achieved, but the number of grid sections increases significantly when obstacles or speed zones are present, leading to increased memory usage and processing time
Solution Approach 1:
The patent segments the geographic area into variable-sized polygons instead of using a fixed grid structure. Each polygon is dynamically sized and positioned to cover specific addresses, allowing the system to represent only the necessary geographic areas rather than an entire fixed grid. This segmentation approach reduces the number of regions from potentially thousands of grid sections to a manageable number of address-specific polygons.
Solution Approach 2:
The patent applies local quality by making each polygon's size and shape specific to its local geographic context. Each polygon is customized to cover its associated addresses and account for local obstacles or speed zones, rather than applying a uniform grid structure. This allows accurate travel time prediction for each location while reducing overall system complexity.
2Productivity
If the number of grid sections is reduced to decrease processing time, then scheduling speed improves, but travel time prediction accuracy decreases due to larger coverage areas per grid section
Solution Approach 1:
The patent implements dynamics by making polygon sizes variable rather than fixed. Each polygon's dimensions are dynamically adjusted based on its specific geographic context, including the distribution of addresses, presence of obstacles, and speed zone boundaries. This dynamic sizing allows the system to maintain high prediction accuracy while using fewer polygons overall, thus improving processing speed without sacrificing precision.
3Ease of operation
If fixed grid size is used for all areas, then system simplicity is maintained, but accuracy decreases in areas with obstacles or varying speed zones where smaller grids are needed
Solution Approach 1:
The patent applies parameter changes by varying the size parameter of geographic regions based on local conditions. Instead of using a fixed grid size throughout, the system adjusts each polygon's dimensions according to factors such as address density, obstacle presence, and speed zone boundaries. This allows the system to maintain simplicity in operation while achieving high accuracy in challenging geographic areas.
Data Source
AI summary
A system for representing predicted travel times between pick-up and drop-off addresses for which routes are to be scheduled. In some embodiments, a computer system defines a disk representing a geographic area surrounding each address having a radius that is selected such that a travel time between any two locations in the area of the disk is within a predetermined time limit. If the travel time from an address to a man-made or natural obstacle that limits travel in the area of the disk is within the predetermined time limit then the radius of the disk is reduced to the distance to the obstacle. A memory stores predicted travel times between addresses encompassed by the geographic areas of different disks.