Vehicle Route Search Using Well-Used Road ETA Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems fail to distinguish between well-used and typical roads, leading to imprecise estimated-time-of-arrival (ETA) calculations and suboptimal route selection, which affects departure and appointment times.
Innovation Solution
An apparatus and method that utilize a GPS module, storage module, and processor to analyze traffic distribution and ETA prediction models to select and modify costs for well-used roads, incorporating these into a priority queue for precise route computation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a cost is computed without distinguishing between well-used roads and typical roads, then the route search process is simple, but the ETA precision deteriorates
Solution Approach 1:
The patent applies local quality by differentiating the cost computation method for well-used roads versus typical roads. Specifically, well-used roads use a deep learning model that dynamically adjusts costs based on time, day, and traffic conditions, while typical roads use standard static cost values. This localized differentiation improves ETA precision without requiring the entire route search system to be complex.
Solution Approach 2:
The patent changes the cost parameter dynamically based on the road type and conditions. For well-used roads, the cost parameter is adjusted according to time of day, day of week, and traffic conditions through the deep learning model. This parameter change enables more accurate ETA calculation for roads where traffic patterns significantly impact travel time.
2Productivity
If multiple vehicles use a well-used road, then the road becomes congested and travel time increases, but the road remains attractive for route selection
Solution Approach 1:
The patent implements feedback by using the deep learning model to continuously update cost values for well-used roads based on current and historical traffic conditions. When multiple vehicles use a well-used road causing congestion, the model detects increased travel times and adjusts the cost parameter upward, providing feedback that influences future route selections to avoid congested periods or alternative paths.
Solution Approach 2:
The patent applies dynamics by making the cost parameter for well-used roads time-dependent and condition-dependent rather than static. The deep learning model adjusts costs dynamically based on time of day, day of week, and observed traffic conditions, enabling the route search system to adapt to changing congestion patterns and select optimal routes under different conditions.
Data Source
AI summary
An apparatus for searching for an optimal route based on a well-used road includes a GPS module configured to detect a current vehicular position. The apparatus also includes a storage module configured to store digital map data. The apparatus also includes a processor configured to select an extension link, which is to be extended according to priority, from among a plurality of search links starting from a departure link, in order to search for an optimal route based on a route search engine on the basis of the current vehicular position and the digital map data. The processor is also configured to work out an ETA cost for the extension link for a search route through an estimated-time-of-arrival (ETA) prediction model based on a well-used road to search for the optimal route.