Neural Network Model Storage Partitioning for Automotive Predictive Maintenance

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional automotive maintenance schedules are often inconvenient and fail to predict component failures in real-time, leading to potential safety hazards and unscheduled vehicle breakdowns.

Innovation Solution

Implementing a system that uses artificial neural networks, specifically spiking neural networks, to analyze sensor data from vehicles to predict maintenance needs, allowing for proactive scheduling of maintenance services based on personalized operating habits and environmental conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional predetermined maintenance schedules are used, then maintenance services can be scheduled regularly, but the system cannot predict component failures in real-time leading to safety hazards and inconvenient timing

Engineering Contradiction:

Improvecomponent failure prediction accuracyVSAvoidmaintenance scheduling convenience

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The system performs preliminary analysis of sensor data using neural networks to predict component failures before they actually occur. This allows maintenance to be scheduled proactively at convenient times rather than reactively after failures, resolving the contradiction between reliability and ease of operation

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The maintenance schedule transitions from static predetermined intervals to dynamic condition-based scheduling. The system continuously monitors sensor data and adjusts maintenance timing based on real-time component health status, enabling both accurate failure prediction and convenient scheduling

Inventive Principle:

Principle #15Dynamics

2Speed

If neural network models are stored in volatile memory, then fast access is achieved, but data loss occurs when power is interrupted

Engineering Contradiction:

Improvemodel access speedVSAvoiddata persistence

Core Design Contradiction:

SpeedVSReliability

Solution Approach 1:

The system merges volatile memory (for fast access) and non-volatile memory (for data persistence) into a unified storage architecture. Neural network models are stored in both memory types simultaneously, allowing the system to leverage the speed of volatile memory while ensuring data persistence through non-volatile memory backup

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The system creates copies of neural network models in both volatile and non-volatile memory. This copying strategy enables fast access operations to use the volatile memory copy while the non-volatile memory copy ensures data persistence and recovery after power interruptions

Inventive Principle:

Principle #26Copying

Data Source

PatentUS11586194B2Storage and access of neural network models of automotive predictive maintenance

Publication Date: 2023.02.21 MICRON TECHNOLOGY INC

AI summary

Systems, methods and apparatus of optimizing neural network computations of predictive maintenance of vehicles. For example, a data storage device of a vehicle includes: a host interface configured to receive a sensor data stream from at least one sensor configured on the vehicle; at least one storage media component having a non-volatile memory; and a controller. The non-volatile memory is configured into multiple partitions (e.g., namespaces) having different sets of memory operation settings configured for different types of data related to an artificial neural network (ANN). The partitions include a model partition configured to store model data of the ANN. The sensor data stream is applied in the ANN to predict a maintenance service of the vehicle. The memory units of the model partition can be configured for read, infrequent updates, improved storage capacity, and/or for access in parallel with input/output for the ANN.