Remote Diagnostic System Using Device Data Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional diagnostic systems for multi-function devices (MFDs) face challenges in efficiently diagnosing and repairing complex issues due to noisy and uncertain device data, leading to time-consuming and costly customer support processes.
Innovation Solution
A remote diagnostic system utilizing a device data classification algorithm that constructs a conditional probability lookup table and score function to map fault codes to service call categories, coupled with a rules engine to analyze diagnostic data and automatically dispatch solutions to customers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional diagnostic systems use device data (fault codes, status codes, usage counters, sensor readings) to remotely diagnose problems, then remote diagnosis capability is provided, but the device data is noisy and uncertain leading to poor classification accuracy
Solution Approach 1:
The system performs preliminary classification of device data into problem types before actual diagnosis. By pre-processing and categorizing the noisy device data into distinct problem types using classification algorithms, the system prepares structured information that improves subsequent diagnosis reliability and reduces the impact of data noise.
Solution Approach 2:
The patent introduces an intermediary classification layer between raw device data and final diagnosis. This intermediary problem type classification acts as a mediator that transforms noisy, uncertain device data into structured problem categories, which then feed into the diagnosis system, thereby improving both measurement precision and reliability.
2Productivity
If conventional systems provide single response to customer calls after analyzing device data, then remote support is provided, but the process is time-consuming and costly for both customer and enterprise
Solution Approach 1:
The system performs preliminary classification of device data into problem types before generating responses. This pre-processing step automatically categorizes issues and prepares structured diagnostic information, enabling faster response generation and reducing the time required for customer support interactions.
Solution Approach 2:
The system enables self-service by automatically classifying device data and generating appropriate responses without requiring extensive manual analysis. The automated classification and response generation capabilities allow the system to serve customers independently, reducing support time and costs for both customers and enterprises.
3Ease of operation
If help-desk calls are used for problem diagnosis, then customer support is provided, but much time is spent gathering preliminary information before addressing the essence of the problem
Solution Approach 1:
The system performs preliminary classification of device data automatically before the customer even contacts support. By pre-categorizing problems and preparing diagnostic information in advance, the system eliminates the need for time-consuming preliminary information gathering during actual support calls, making problem diagnosis easier and faster.
4Extent of automation
If device data is transmitted to remote server for analysis, then remote diagnosis is enabled, but the classification of device data with respect to service action succeeds with only a small fraction of calls
Solution Approach 1:
The system performs preliminary classification into problem types as a first step in automation. By pre-categorizing device data into distinct problem types using classification algorithms, the system establishes a foundation that improves the accuracy of subsequent service action classification, thereby increasing the fraction of calls where automated classification succeeds.
Data Source
AI summary
A remote diagnostic system and method based on device data classification. Device diagnostic data with respect to a device can be acquired and a conditional probability look up table can be constructed for each fault code associated with the device diagnostic data by a classification module. A score function can then be created by summing the conditional probabilities and an occurrence of the fault code can be mapped to a service call category with a numerically highest score function. The fault occurrence data in association with a number of time stamps and device identifiers can be stored in a data warehouse. The occurrence of fault code can be matched with respect to a solution set which can be automatically dispatched to a customer via a communications link.