Stochastic Routing Policy for On-Time Arrival Probability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems fail to provide routes that maximize the probability of arriving at a destination within a specific time budget, as they rely on deterministic algorithms that do not account for unpredictable delays, leading to potential late arrivals.
Innovation Solution
A method using stochastic routing algorithms that set a time budget based on the shortest expected journey time multiplied by a scaling factor, allowing the system to dynamically adjust routes based on actual travel times to maximize the probability of on-time arrival, with the scaling factor adjustable by user input or predetermined values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If deterministic routing algorithms are used to calculate the fastest route, then the route computation speed is improved, but the reliability of on-time arrival deteriorates due to unpredictable delays
Solution Approach 1:
The patent applies dynamics by transitioning from static deterministic routing to dynamic stochastic routing. The system uses probability distributions that are updated in real-time based on live traffic data, allowing the routing algorithm to adapt its cost calculations dynamically. This enables the system to maintain computational efficiency while improving on-time arrival reliability by adjusting route recommendations based on current conditions and uncertainty modeling.
Solution Approach 2:
The patent changes the fundamental parameter representation from fixed deterministic costs to probabilistic cost distributions. Each road segment is assigned a probability distribution of travel times rather than a single expected value, allowing the algorithm to account for variability and uncertainty. This parameter transformation enables the routing system to optimize for reliability while maintaining computational tractability through efficient probability distribution management.
2Reliability
If stochastic routing algorithms are used to maximize on-time arrival probability, then the on-time arrival reliability is improved, but the computational complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the routing problem into independent road segment components, each with its own probability distribution. Instead of computing over the entire route at once, the system processes individual segments separately, combining their probability distributions through convolution or approximation methods. This segmentation reduces computational complexity while maintaining the ability to calculate overall on-time arrival probability for the complete route.
Solution Approach 2:
The patent uses copying by pre-computing and storing probability distribution characteristics for road segments based on historical data, then reusing these pre-computed distributions for real-time routing decisions. This approach avoids repeated complex calculations by copying previously analyzed probability patterns, significantly reducing computational complexity during actual route planning while maintaining accurate stochastic modeling.
3Measurement precision
If time-dependent average speed data is used to generate accurate routes, then the route accuracy is improved, but the data processing requirements increase
Solution Approach 1:
The patent applies preliminary action by pre-processing historical traffic data to extract and store probability distribution parameters for each road segment before real-time routing is needed. This pre-computation includes calculating mean, variance, and other statistical characteristics from historical speed data, so that during actual route planning, the system only needs to combine these pre-computed parameters rather than processing raw historical data again. This significantly reduces data processing requirements while maintaining high route accuracy.
Solution Approach 2:
The patent transforms raw time-dependent speed data into simplified probability distribution parameters (mean, variance, shape parameters). Instead of storing and processing complete time-series speed data for every road segment, the system converts this detailed data into compact parametric representations that capture the essential variability. This parameter transformation reduces data volume and processing requirements while preserving the information needed for accurate stochastic routing.
Data Source
Figure 1~3
Figure 2
Figure 4~5
AI summary
A method of generating a routing policy from an origin to a destination through a navigable network, the navigable network being represented by an electronic map comprising a plurality of segments representing navigable segments of the navigable network, each segment having associated therewith data indicative of an expected travel time along the segment. The expected travel times are used for respective segments of the electronic map to determine a route from the origin to the destination that has a shortest expected journey time using a deterministic routing algorithm. A stochastic routing algorithm is then used to generate a routing policy that maximises the probability of arriving at the destination within a time budget from leaving the origin, wherein the time budget is based at least on said determined shortest expected journey time. A route from the origin to the destination can then be generated by applying the routing policy to a scenario in which each segment is transited in a time based on the data indicative of the expected travel time associated with the respective segment.