Locator Layout Optimization for Fixture Cost and Part Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge lies in determining the optimal number and positioning of locators on a vehicle fixture to balance fixture cost and product quality, as fewer locators may compromise quality while more locators increase costs.
Innovation Solution
A computer system employing a trained artificial neural network collects part parameters and input data sets to predict the target number and distance of locators, optimizing for minimal deviation and deflection, thereby determining the optimal configuration for good quality and minimal cost.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If more locators are used to fix the part on the fixture, then product quality and stability are improved, but fixture cost increases
Solution Approach 1:
The patent uses artificial neural networks to optimize key parameters including the number of locators, their positions, and spacing distances. By computationally determining optimal parameter values based on part geometry and material properties, the system achieves high manufacturing precision with minimal locators, resolving the contradiction between quality and cost.
Solution Approach 2:
The patent replaces traditional mechanical trial-and-error methods for determining locator configuration with an artificial intelligence-based computational system. The neural network model predicts optimal locator arrangements by processing part parameters and geometric data, substituting physical experimentation with digital simulation and optimization.
2Device complexity
If fewer locators are used to reduce fixture cost, then device complexity is reduced, but product quality and stability deteriorate
Solution Approach 1:
The system optimizes locator parameters (number, position, spacing) to find the minimum configuration that still achieves required quality thresholds. The neural network evaluates multiple parameter combinations to identify the optimal point where quality requirements are met with the fewest locators possible.
Solution Approach 2:
The patent applies the principle of partial action by determining the minimum necessary number of locators required to achieve acceptable quality levels. Rather than using excessive locators for all cases, the system calculates the precise minimum configuration needed for each specific part, avoiding unnecessary complexity while maintaining quality.
3Device complexity
If the distance between locators is increased, then fixture cost is reduced, but part deviation and deflection increase
Solution Approach 1:
The neural network optimizes the spacing distance parameter between locators by evaluating its impact on part deviation and deflection. The system determines optimal spacing values that minimize measurement errors while reducing the total number of locators required, balancing cost and precision.
4Device complexity
If the distance between locators is increased, then fixture cost is reduced, but part form accuracy deteriorates
Solution Approach 1:
The system optimizes locator spacing parameters to find the optimal balance between cost and form accuracy. The neural network evaluates how spacing distances affect part form measurement accuracy and determines maximum acceptable spacing values that maintain quality requirements while minimizing the number of locators.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A computer system (100) comprising a processor device (102) implementing a trained artificial neural network (ANN) for determining a target number of locators and a target distance between adjacent locators placed on a part to fix the part on a fixture, the computer system (100) being configured to: - collect a part (201) associated with part parameters (PP1, PP2, PP3); - collect a first set (202) of input dublets (ID1, ID2), each input dublet (ID1, ID2) comprising a first value (ID1) and a second value (ID2), the first value (ID1) being a number of locators and the second value (ID2) being a distance between adjacent locators; - for each input dublet of the first set, collect as input (203) to the trained artificial neural network the part parameters and said input dublet of the first set; - for each input dublet of the first set, provide (204) an output dublet (OD1, OD2) predicted by the trained artificial neural network (ANN), the output dublet (OD1, OD2) comprising a first output value (OD1) and a second output value (OD2), the first output value (OD1) being a part deviation defined as a change of a part dimension from a nominal part dimension, and the second output value (OD2) being a part deflection defined as a change of a part form from a nominal part form, so that a second set of output dublets (OD1, OD2) is collected, each output dublet (OD1, OD2) of the second set resulting from an input dublet (ID1, ID2) of the first set; - select (205) a selected output dublet (OD1, OD2) among the output dublets (OD1, OD2) of the second set, the selection (205) being based on an optimisation criteria, said selected output dublet (OD1, OD2) resulting from a determined input dublet (ID1, ID2) of the first set; - determine (206) the target number of locators as the number of locators (ID1)of the determined input dublet (ID1, ID2), and determine (206bis) the target distance between adjacent locators as the distance between adjacent locators (ID2) of the determined input dublet (ID1, ID2).