Dynamic Route Planning Using Stochastic Time-Dependent Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional route planners for public transportation fail to provide accurate travel times due to their reliance on static schedules, ignoring real-time traffic conditions and peak/off-peak variations, leading to user dissatisfaction.
Innovation Solution
A computer-implemented dynamic route planning system that uses stochastic time-dependent models based on travel smart card data to generate continuous distributions of travel and waiting times, associating these with a transportation graph and applying a modified multi-criteria shortest path technique to optimize route recommendations, considering the FIFO property and component aggregation errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If static schedules are used for route planning, then the system is simple to implement, but the travel time estimation accuracy deteriorates
Solution Approach 1:
The patent transforms the static route planning system into a dynamic one by incorporating real-time traffic conditions and time-dependent travel time estimates. The system now adapts route recommendations based on current traffic situations, making the planning process dynamic rather than fixed, which directly addresses the accuracy problem while managing complexity through structured data collection and processing.
Solution Approach 2:
The system implements feedback mechanisms by continuously collecting actual travel time data from commuters and using this information to update and refine travel time distributions. This feedback loop allows the system to learn from real-world performance and improve its estimates over time, resolving the contradiction between system simplicity and estimation accuracy.
2Ease of operation
If static route planners are used, then the system is easy to operate, but the results are inadequate for real-world traffic conditions
Solution Approach 1:
The system employs self-service principles by automatically collecting travel data from commuters via smart cards and using this data to generate and update travel time distributions without manual intervention. The system serves itself by continuously improving its own accuracy through automated data collection and processing, maintaining ease of operation while enhancing reliability.
Solution Approach 2:
The patent changes key parameters from fixed static values to dynamic time-dependent distributions. Travel times are no longer constant but are represented as probability distributions that vary with time of day, day of week, and traffic conditions. This parameter transformation allows the system to provide more reliable recommendations while maintaining user-friendly operation through automated processing.
3Measurement precision
If travel time data is collected and processed to generate continuous distributions, then the accuracy of travel time estimation is improved, but the data processing complexity increases
Solution Approach 1:
The patent segments the travel time data collection and processing into distinct components: data collection from smart cards, data cleaning and validation, distribution parameter estimation, and route calculation. This segmentation allows each component to be optimized independently, improving measurement accuracy while managing overall system complexity through modular processing steps.
Solution Approach 2:
The system performs preliminary actions by pre-processing and storing travel time data in organized structures before actual route planning queries. Travel time distributions are pre-calculated and stored for various origin-destination pairs and time periods, allowing fast query response without real-time computation, thus improving accuracy while controlling processing complexity.
4Measurement precision
If modified multi criteria shortest path technique is used, then the route recommendation accuracy is improved, but the computational complexity increases
Solution Approach 1:
The patent transforms the traditional shortest path problem by changing the cost parameter from a single fixed value to a multi-criteria evaluation based on expected travel time, variance, and other factors. The modified algorithm incorporates these multiple parameters to provide more accurate route recommendations, accepting increased computational complexity as necessary for improved accuracy in stochastic time-dependent networks.
Data Source
AI summary
A computer-implemented technology for planning routes is described herein. In accordance with one aspect, travel data of commuters of a transportation network are provided. Continuous distributions of travel time and waiting time are generated from the travel data. The continuous distributions of travel time and waiting time are associated to a transportation graph of the transportation network. The transportation graph includes nodes corresponding to stops of the transportation network and edges interconnecting the nodes. Travel time and waiting time are associated as costs of the edges in the transportation graph. In response to receiving input parameters, expected costs of candidate routes in the transportation graph are determined in accordance with a modified multi criteria shortest path technique. The modified multi criteria shortest path technique invokes a subroutine to retrieve accurate costs of routes based at least on the costs of the edges in the transportation graph. Route recommendations are provided based on the expected costs of the candidate routes.