Dynamic POI Duration Prediction Using Historical Time Functions

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing systems for obtaining and using data related to dynamic points of interest (POIs), such as speed limit enforcement devices, face challenges in accurately determining the duration time of these POIs, as current techniques often assign fixed durations based on attributes like country or road class, which may not reflect actual operational times.

Innovation Solution

The method involves obtaining a duration time function that indicates how duration time varies with respect to time and location, based on historical data from multiple dynamic POIs. This function is then associated with geographic location data, allowing for more accurate duration time predictions for new dynamic POIs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If fixed duration times are assigned to dynamic POIs based on attributes like country or road class, then the system is simple to operate and implement, but the accuracy and freshness of the duration time data deteriorates

Engineering Contradiction:
Improveduration time accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies dynamics by transitioning from static fixed duration assignments to dynamic duration predictions. The system uses historical data and machine learning models to continuously update and adapt duration predictions based on changing conditions, time patterns, and location-specific characteristics, making the duration information dynamic rather than fixed

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent applies preliminary action by pre-processing historical POI data to train machine learning models before actual use. The system performs offline training and model generation in advance, so that when new POIs are encountered, the system can immediately apply pre-trained models to predict durations without real-time computation delays

Inventive Principle:
Principle #10Preliminary action

Solution Approach 3:

The patent introduces machine learning models as intermediaries between raw historical data and duration predictions. These models act as mediators that learn complex patterns from historical data and translate them into accurate duration predictions, bridging the gap between simple attribute-based assignment and complex real-time analysis

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If duration time data is continuously updated based on real-time user reports, then the freshness and accuracy of POI information improves, but the loss of time and computational resources increases

Engineering Contradiction:
ImprovePOI data reliabilityVSAvoiddata processing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary action by pre-training machine learning models offline using historical data before deployment. This advance preparation allows the system to make rapid predictions without real-time computation, reducing processing time while maintaining high reliability through pre-learned patterns from extensive historical data

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms where user reports and actual POI observations are continuously fed back into the model training process. This feedback loop allows the system to learn from real-world data, continuously improving prediction accuracy while efficiently processing new information through already-trained models

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3872455B1Methods of obtaining and using point of interest data
Publication Date: 2025.02.12 TOMTOM TRAFFIC
  • EP3872455B1 patent drawingFigure 1~2
  • EP3872455B1 patent drawingFigure 3
  • EP3872455B1 patent drawingFigure 4~7

AI summary

Data identifying a geographic location of a dynamic POI is obtained. A database storing data indicative of a plurality of duration time functions is accessed, each duration time function being associated with data indicative of a geographic location, and each duration time function being indicative of a variation in duration time with respect to time. The duration time for a dynamic POI is the time that the dynamic POI remains in a first state before changing to a second state following initiation of the first state. Data indicative of the duration time for the dynamic POI is obtained by identifying an applicable duration time function from the database based upon at least the geographic location of the POI, and the identified duration time function is used to determine a duration time for the dynamic POI for a time of interest.