Rule-Based Decision Filtering for Complex Automated Decisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional rule-based systems for complex automated decision making, such as in autonomous vehicles, become overly complex, resource-intensive, and brittle, while machine learning approaches can make poor decisions in unfamiliar situations, lacking clear rules for object identification from images.
Innovation Solution
Implement a generate-filter-select decision process incorporating commutative action rule-based systems (CARBS) and automatic root cause analysis (ARCA) using ternary fault scenarios to optimize decision making and fault detection, reducing the need for conflict resolution and reliance on statistical correlations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional rule-based systems are used for complex automated decision making, then decision-making accuracy is maintained through clear rules, but system complexity and resource consumption increase significantly
Solution Approach 1:
The patent segments the complex rule-based system into modular components: a rule engine that processes individual rules, a conflict resolution module that handles rule interactions, and a decision execution unit. This segmentation allows the system to manage complexity through organized modularity while maintaining the reliability of rule-based decision-making.
Solution Approach 2:
The patent introduces an intermediary conflict resolution layer between rule processing and decision execution. This mediator resolves conflicts among multiple rules before final decisions are made, enabling the system to handle complex rule interactions without requiring complete rewriting of the rule base, thus maintaining accuracy while managing complexity.
2Reliability
If traditional rule-based systems are used for complex automated decision making, then decision-making reliability is maintained, but processing time and resource intensity increase
Solution Approach 1:
The patent implements preliminary action by pre-processing and categorizing rules into groups based on their decision domains and conflict potentials. This pre-organization allows the rule engine to quickly retrieve and evaluate only relevant rules during execution, reducing processing time while maintaining the comprehensive coverage needed for reliable decision-making.
Solution Approach 2:
The patent employs periodic action through iterative rule evaluation cycles with configurable priorities. Instead of evaluating all rules simultaneously, the system processes rules in periodic batches with different priority levels, allowing critical rules to be evaluated first while less urgent rules are processed subsequently, thereby reducing overall processing time without compromising decision reliability.
3Adaptability or versatility
If machine learning approaches are used for object identification, then adaptability to new situations improves, but decision-making accuracy deteriorates in unfamiliar scenarios due to lack of clear rules
Solution Approach 1:
The patent merges machine learning components with traditional rule-based systems into a hybrid architecture. Machine learning models handle pattern recognition and object identification tasks where adaptability is needed, while rule-based modules provide deterministic decision logic for safety-critical and unfamiliar scenarios. This combination allows the system to leverage the strengths of both approaches, maintaining accuracy where rules apply while gaining adaptability where learning excels.
Data Source
AI summary
A set of input conditions is obtained. A plurality of potential decisions is obtained based at least in part on the set of input conditions. A rule-based system is used to process the plurality of potential decisions and obtain a set of one or more updated potential decisions, wherein: the rule-based system specifies a plurality of rules; a rule specifies a rule condition and a corresponding action, wherein when the rule condition is met, the corresponding action is to be performed; and using the rule-based system to process the plurality of potential decisions includes: for a selected potential decision in the plurality of potential decisions, determining whether the rule condition is met for a selected rule among the plurality of rules, wherein the selected rule condition is dependent on, at least in part, the selected potential decision; and in response to the selected rule condition being met, performing the corresponding action. The set of one or more updated potential decisions to be executed is output.