Personalized Alert Ranking for Medical Device Maintenance Queues
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to provide personalized and efficient distribution and ordering of maintenance alerts to remote service engineers (RSEs), leading to suboptimal alert handling and increased downtime of medical devices.
Innovation Solution
A system that utilizes an alert ranking machine learning (ML) model trained on historical maintenance alerts data to generate personalized ranked lists of unresolved alerts for individual RSEs, considering their expertise, experience, and other alert characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If alerts are simply ordered by priority for all RSEs, then the alert distribution system is simple and easy to operate, but it fails to account for individual RSE expertise and experience, leading to suboptimal alert handling
Solution Approach 1:
The system pre-trains machine learning models on historical alert data and RSE performance data before deployment. This preliminary action enables the system to automatically generate personalized alert rankings for each RSE based on their expertise and experience, eliminating the need for manual customization while optimizing alert handling efficiency from the start
Solution Approach 2:
The system dynamically changes the ranking parameters of alerts based on individual RSE characteristics such as expertise areas, experience level, and historical performance. Instead of using a fixed priority ordering, the system adjusts alert rankings by incorporating RSE-specific parameters, thereby optimizing alert distribution for each engineer's capabilities
2Reliability
If a large number of alerts are generated from multiple predictive models, then comprehensive monitoring is achieved, but identifying high-priority alerts that match RSE expertise becomes increasingly difficult
Solution Approach 1:
The system introduces machine learning models as intermediary components between the predictive models and RSEs. These models process the large volume of alerts from multiple predictive models, filter them based on RSE expertise and experience, and present a personalized ranked list to each RSE, thereby simplifying alert identification without compromising monitoring coverage
Solution Approach 2:
The system segments the large set of alerts into personalized subsets for each RSE based on their expertise areas and experience levels. By dividing the monolithic alert queue into tailored segments, each RSE receives only the most relevant high-priority alerts, reducing the complexity of identifying actionable items while maintaining comprehensive monitoring across all systems
3Productivity
If RSEs review alerts in top-down fashion by priority, then high-priority alerts are addressed first, but important alerts may be overlooked if they require specialized knowledge not possessed by the reviewing RSE
Solution Approach 1:
The system applies local quality by customizing the alert ranking for each RSE based on their specific expertise, experience, and performance history. Instead of a uniform priority ordering for all RSEs, each engineer receives a personalized alert list that prioritizes alerts matching their local knowledge and capabilities, thereby improving both response speed and resolution accuracy
Data Source
AI summary
In an alerting system, one or more predictive models are trained to generate maintenance alerts for medical devices of a fleet of medical devices based on machine log data received from the medical devices. Historical maintenance alerts data are stored including at least historical maintenance alerts generated by the one or more predictive models for the fleet of medical devices. Instructions are readable and executable by at least one electronic processor to: train an alert ranking machine learning (ML) model to rank alerts of a queue of alerts using the historical maintenance alerts data; receive unresolved alerts for medical devices of the fleet from the one or more predictive models; generate a ranked list of the unresolved alerts allocated to a service engineer (SE) using the trained ranking ML model; and provide, on a display device accessible by the SE, the ranked list of the unresolved alerts allocated to the SE.