Rule-Based Perception Augmentation for Noisy Machine Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Machine vision systems face challenges in accurately inferring discrete semantic states due to noisy sensor data and occlusions, making it difficult to maintain and understand custom-designed program code for post-processing symbolic outputs.
Innovation Solution
A logical-perception-augmentation system that uses domain-specific logical rules, including augmentation and persistence rules, to infer and persist the states of physical system components, enhancing perception beyond sensor data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If machine vision equipment captures images to infer system states, then visual information is obtained, but the captured information is noisy and intermittent
Solution Approach 1:
The patent introduces logical rules as an intermediary between raw sensor data and system state determination. These rules process and interpret the noisy visual information, filtering out errors and filling in gaps when components are occluded or temporarily out of view, thereby transforming unreliable raw data into reliable state perceptions.
Solution Approach 2:
The system performs preliminary actions by establishing logical rules in advance that define expected relationships and states. When sensor data is noisy or incomplete, these pre-established rules can immediately apply reasoning to determine correct states without waiting for additional sensor input, thus maintaining reliable perception despite intermittent data quality issues.
2Measurement precision
If custom-designed program code is used for post-processing symbolic outputs, then state inference is achieved, but the code is difficult to maintain and understand
Solution Approach 1:
The patent replaces complex custom-designed program code with a rule-based system that copies the structure of human expert knowledge. Instead of maintaining intricate algorithms, the system uses declarative logical rules that mirror domain expertise, making the system easier to understand, validate, and maintain while preserving state inference accuracy.
Solution Approach 2:
The patent substitutes the mechanical system of custom program code with a rule-based logical reasoning system. This replacement transforms the approach from procedural manipulation of symbolic outputs to declarative logical inference, significantly reducing code complexity and improving maintainability while maintaining or enhancing inference precision.
3Reliability
If logical rules are used to augment perception, then accurate state inference is achieved, but system complexity increases
Solution Approach 1:
The patent implements a universal rule-based framework that handles multiple functions: filtering noisy data, inferring occluded states, maintaining persistence across time steps, and coordinating multiple sensors. This single rule-based system replaces what would otherwise require multiple separate processing modules, managing complexity through universality rather than proliferation of components.
Data Source
AI summary
One embodiment can provide a system for augmenting perceptions of a machine sensing system. During operation, one or more sensors of the system can obtain sensory information associated with a physical system. The system can determine a state of a first component of the physical system based on the obtained sensory information; select, from a rule database, one or more logical rules associated with at least the first component; and augment a perception of the machine sensing system toward the physical system based on both the determined state of the first component and the selected one or more logical rules, thereby facilitating the machine sensing system to make a decision associated with the physical system.