Vehicle ML Update Loop From User Feedback Clusters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Vehicle applications and machine learning models are not timely updated to address new issues, leading to user discomfort due to inadequate handling of challenges, weaknesses, or edge cases, such as excessive deceleration upon detecting vehicles ahead.
Innovation Solution
A system that automatically generates requirements for vehicle applications and machine learning models based on user feedback, using a requirement defining machine-learning model to continuously update the models for user experience improvement without manual developer input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual developer input is used to update vehicle applications and machine learning models, then updates can be made with detailed understanding of requirements, but the updating process is slow and not timely
Solution Approach 1:
The system enables self-service by allowing vehicle applications and machine learning models to automatically generate their own requirements and perform updates without manual developer intervention. The requirement defining machine learning model autonomously creates requirements from user feedback, and the system automatically trains and deploys updated models, eliminating the need for continuous manual developer involvement while maintaining timely updates.
Solution Approach 2:
The system implements a closed-loop feedback mechanism where user feedback is continuously collected and automatically processed to generate new requirements. This feedback loop enables the system to learn from actual user experiences and continuously improve the machine learning models in real-time, ensuring timely updates that directly address user needs without manual intervention.
2Productivity
If automatic requirement generation is implemented, then continuous and timely updates are enabled, but the complexity of the system increases
Solution Approach 1:
The system replaces the mechanical process of manual requirement writing and model updating with an automated machine learning-based system. Instead of developers manually analyzing user feedback and writing requirements, a requirement defining machine learning model automatically generates requirements from user feedback. This substitution of manual mechanical processes with automated intelligent systems enables continuous updates while managing complexity through automation.
Solution Approach 2:
The requirement defining machine learning model serves multiple functions: it processes user feedback, generates requirements, and guides model training. This multi-functional component consolidates what would otherwise require separate manual processes, enabling automatic requirement generation and continuous updates while reducing overall system complexity through functional integration.
3Reliability
If continuous updating is performed to address new issues, then user experience is improved, but the manual effort and resources required increase
Solution Approach 1:
The system enables self-service by automatically generating requirements and updating models without requiring manual developer effort for each update. The requirement defining machine learning model autonomously processes user feedback and creates requirements, and the system automatically trains and deploys updated models, making the continuous updating process easy to maintain without increasing manual resources.
Solution Approach 2:
The system performs preliminary action by continuously monitoring user feedback and automatically preparing requirement updates before issues become problematic. The requirement defining machine learning model proactively generates requirements from incoming user feedback, and models are trained and deployed in advance, enabling continuous improvement without requiring reactive manual intervention for each new issue.
Data Source
AI summary
Continuous update of driving system for user experience improvement is performed by collecting a plurality of user feedback from an Internet, the plurality of user feedback identified by an identification machine-learning model to involve one or more vehicles, clustering, by a clustering machine-learning model, the plurality of user feedback into a plurality of incident clusters, and defining, by a requirement defining machine-learning model, a user experience requirement according to user feedback among the plurality of incident clusters.