Image Reproduction Job Duration Estimation via Iterative Correction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image reproduction systems face inaccuracies in predicting process durations for print jobs due to limited accuracy in estimated durations and variations caused by environmental changes and action combinations, leading to significant prediction errors, especially in mixed-type print queues.

Innovation Solution

An iterative process updates correction factors for each job type based on actual processing times, applying these updates to all similar jobs in the queue, allowing for precise prediction of process durations across different job types by continuously refining correction factors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If model-based duration estimates are used for scheduling print jobs, then scheduling can be performed, but the prediction accuracy is limited due to environmental variations and action combinations

Engineering Contradiction:
Improvescheduling capabilityVSAvoidprocess duration prediction accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent implements a feedback mechanism where actual processing times from completed jobs are continuously fed back to update correction factors. The controller compares model-based estimates with actual durations, calculates correction factors, and applies these to improve future predictions for similar job types, thereby resolving the accuracy limitation while maintaining scheduling capability

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent dynamically adjusts prediction parameters by introducing correction factors that modify the base model estimates. These correction factors are derived from actual measurement data and are updated over time to adapt to environmental changes and system variations, transforming static model parameters into adaptive predictors

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If correction factors are updated iteratively over all jobs of the same type, then prediction accuracy improves asymptotically, but computational complexity increases

Engineering Contradiction:
Improveprocess duration prediction accuracyVSAvoiditerative update process complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies correction factors locally to specific job types rather than globally to all jobs. Each job type maintains its own correction factor based on its specific characteristics, allowing accurate predictions for diverse job types without requiring complex global models. This local approach reduces overall system complexity while maintaining high prediction accuracy

Inventive Principle:
Principle #3Local quality

3Measurement precision

If separate correction factors are stored for multiple job types, then accuracy improves for different job types, but memory requirements and data management complexity increase

Engineering Contradiction:
Improveprediction accuracy for different job typesVSAvoidnumber of correction factors to store
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent creates a universal correction factor update mechanism that handles multiple job types through a common process. The same iterative update algorithm and data structures are used for all job types, allowing the system to efficiently manage correction factors for numerous job types without proportionally increasing complexity. The system universally applies the correction factor methodology across different job types

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3196753B1Method of estimating a process duration for an image reproduction job
Publication Date: 2018.10.03 OCE HLDG BV
  • EP3196753B1 patent drawingFigure 1~2
  • EP3196753B1 patent drawingFigure 3
  • EP3196753B1 patent drawingFigure 4~5

AI summary

A method of controlling an image reproduction system, wherein a process duration for processing a reproduction job is estimated on the basis of a measured time (D) for the processing of a number (n) of pages that have been processed already, characterized by the steps of: - storing a digital table which contains a number of pre-defined job types as well as a respective correction factor (A4) for each job type; - when a job is received (S1), assigning (S2) a job type from the digital table to the received job; - calculating (S3) a first estimate (D1) for a time needed for processing at least a specified collection of pages from the job, said first estimate being calculated on the basis of a model that reflects at least the dependency of the processing time on the number of pages to be processed; - calculating (S4) a second estimate (D2) by multiplying the first estimate (D1) with the correction factor (A4) that is related to the assigned job type; - when said specified collection of pages has been processed, calculating a deviation of the first estimate (D1) from the processing time (D) that has actually been needed, and updating (S9) the correction factor (A4) on the basis of that deviation; and - storing the updated correction factor in the digital table.