ETO Configuration Solver Using DNN Logic Approximation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current engineer-to-order (ETO) configuration systems force users to follow a strictly predefined order for setting parameters, which is inefficient and impractical, especially when external tools like simulators are involved, leading to significant computation time and combinatorial explosions during the configuration process.
Innovation Solution
An ETO configuration method and system that allow users to set parameters in any order by using a trained deep neural network (DNN) to approximate the behavior of external configuration tools, enabling efficient determination of output configuration values without requiring sequential parameter input, and utilizing a solver to evaluate first-order logic functions for piecewise linear mappings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a strictly predefined order for parameter input is enforced in ETO configuration systems, then the configuration process can be systematically managed, but the user flexibility and efficiency are significantly reduced
Solution Approach 1:
The patent introduces a solver as an intermediary component that mediates between the user's arbitrary parameter inputs and the configuration system. The solver evaluates first-order logic formulas to determine valid parameter combinations, allowing users to input parameters in any order while the solver ensures configuration consistency. This intermediary approach resolves the contradiction by providing user flexibility without compromising system manageability.
Solution Approach 2:
The patent changes the fundamental parameter of parameter input order from fixed to variable. By representing configuration rules as first-order logic formulas with quantifiers, the system allows parameters to be instantiated in any order while maintaining logical consistency through the solver's evaluation process. This parameter change enables user flexibility while the formal logic framework maintains system manageability.
2Reliability
If external tools like simulators are invoked repeatedly during configuration validation, then configuration accuracy is ensured, but computation time increases significantly
Solution Approach 1:
The patent creates a logical copy of the configuration validation process by representing external tool validation requirements as first-order logic formulas. Instead of repeatedly invoking external simulators, the system evaluates these logical formulas once during configuration, then reuses the results. This copying approach maintains validation accuracy while dramatically reducing computation time by avoiding repeated external tool invocations.
Solution Approach 2:
The patent performs preliminary action by encoding all validation rules as first-order logic formulas before the configuration process begins. The solver evaluates these pre-encoded formulas to determine valid parameter combinations, eliminating the need for repeated external tool invocations during configuration. This preliminary encoding and evaluation approach ensures validation accuracy while minimizing computation time.
3Productivity
If all parameters must be provided in a strictly predefined order, then the configuration system can maintain consistency, but the productivity and user efficiency deteriorate
Solution Approach 1:
The patent replaces the mechanical sequential processing system with a logical evaluation system. Instead of enforcing a fixed input order through mechanical process control, the system uses first-order logic formulas and a solver to evaluate parameter consistency regardless of input order. This substitution enables high productivity through arbitrary parameter input while maintaining configuration consistency through logical evaluation.
Solution Approach 2:
The patent creates a universal configuration approach where the solver can handle any parameter input order and any combination of parameters. The first-order logic framework provides multi-functionality, allowing the system to validate configurations regardless of which parameters are provided first or in what sequence. This universality enables both high productivity and consistent configuration outcomes.
Data Source
AI summary
A method for determining a set of output configuration values characterizing a specific configuration of a complex product, includes receiving a set of input configuration parameters, providing at least a part of the input configuration parameters as an input to a solver, using the solver to calculate at least one output value from the provided input configuration parameters, and determining the set of output configuration values from at least the output value calculated by the solver. The solver is configured for solving a first order logic function encoding an algorithm of a trained deep neural network or DNN, wherein the algorithm of the DNN has been trained for modeling a function of an external configuration tool or ECT that is required for determining the specific configuration of the complex product. A system for determining the set of output configuration values, and a training system, are also provided.