AI Apparatus Self-Diagnosis via Shock Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users of artificial intelligence apparatuses face challenges in diagnosing internal defects after a shock, as existing self-diagnosis programs are not frequently used and cannot provide real-time detection of part abnormalities, leading to delayed repairs and exacerbated defects.
Innovation Solution
An AI apparatus equipped with a sensing unit including a gyroscope sensor, acceleration sensor, and proximity sensor, which uses a deep learning model to determine if a shock requires self-diagnosis and performs prioritized self-diagnosis by inputting sensor data, estimating drop angles, and outputting diagnosis results to inform users and service centers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual diagnosis programs are used, then users can perform basic diagnostics, but real-time detection of part abnormalities is not achieved and defects are not detected immediately
Solution Approach 1:
The system performs preliminary self-diagnosis automatically when a shock is detected, before the user can manually diagnose. The processor executes the self-diagnosis program immediately upon detecting abnormal acceleration patterns, identifying defective parts in advance of any user action.
Solution Approach 2:
The system provides immediate feedback to the user about defective parts through the output unit. When the self-diagnosis identifies a defective part, the system automatically notifies the user and provides repair information, creating a closed-loop feedback system that eliminates delayed detection.
2Reliability
If self-diagnosis programs are frequently executed, then real-time defect detection is achieved, but device complexity and power consumption increase
Solution Approach 1:
Instead of continuously executing full self-diagnosis programs, the system performs partial diagnostics only when necessary - specifically when shock patterns are detected by the acceleration sensor. This selective execution maintains real-time detection capability while minimizing computational overhead and device complexity.
Solution Approach 2:
The system changes operational parameters by switching between normal operation mode and self-diagnosis mode based on acceleration sensor input. When shock patterns are detected, the processor transitions to executing self-diagnosis routines; otherwise, it maintains normal operation, thereby reducing overall system complexity while preserving real-time detection capability.
3Measurement precision
If comprehensive sensor data is collected for self-diagnosis, then accuracy of shock classification is improved, but data processing time and computational load increase
Solution Approach 1:
The acceleration sensor continuously monitors for shock patterns in advance, and when a potential shock is detected, the system immediately captures relevant sensor data and initiates self-diagnosis. This preliminary detection approach ensures accurate shock classification without requiring continuous processing of all sensor data.
Solution Approach 2:
The system extracts only the necessary sensor data relevant to shock detection and self-diagnosis, rather than processing all available sensor information. The processor selectively uses acceleration sensor data and other directly relevant inputs, eliminating unnecessary computational overhead while maintaining diagnostic accuracy.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables real-time self-diagnosis and prioritization of defect detection in AI apparatuses, providing timely warnings and improving repair efficiency by determining the severity and location of shocks, thus enhancing user convenience and system reliability.
Implementation Method 1
a sensing unit that includes a gyroscope sensor and an acceleration sensor
Implementation Method 2
a sensing unit that includes a gyroscope sensor and an acceleration sensor
Data Source
AI summary
Disclosed is an artificial intelligence apparatus that includes: a sensing unit that includes a gyroscope sensor and an acceleration sensor; an output unit that outputs information; and a processor that acquires a resultant value output from a deep learning model by inputting data acquired from the gyroscope sensor and the acceleration sensor into the deep learning model, performs self-diagnosis when the resultant value shows a shock requiring self-diagnosis, and controls the output unit to output a result of the self diagnosis, in which the deep learning model is a neural network having an adjusted weight by being trained using whether a shock is a shock requiring the self-diagnosis as a resultant value and sensing values of the gyroscope sensor and the acceleration sensor as input values.