Unified Connected Network with NLP for Multi-Domain Events

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing systems struggle with handling large volumes of data from multiple domains, lack real-time intelligent responses, and are limited to processing data from a single domain, such as telematics, failing to trigger services when pre-programmed responses are absent.

Innovation Solution

A unified connected network system that utilizes a server to receive and analyze data from various electronic devices, employing machine-learning algorithms to determine personalized responses across multiple domains, including automotive, banking, healthcare, and home automation, and transmit these responses to control connected devices.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a server uses pre-programmed responses stored in a mapping for event processing, then the response time is fast and system complexity is low, but the system cannot handle huge volumes of data, cannot generate responses for new event types, and is limited to a single domain

Engineering Contradiction:
Improvedomain coverageVSAvoidsystem complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent introduces a natural language processing intermediary layer between the multi-domain data input and the response generation system. This NLP intermediary converts diverse data from multiple domains into structured queries that can be processed by the server, enabling the system to handle data from banking, medical, automotive, and other domains without requiring separate processing systems for each domain

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent transforms the system from using fixed pre-programmed responses to generating dynamic responses based on natural language processing. By changing the parameter from static mapping tables to dynamic NLP-based query processing, the system gains the ability to handle new event types and multiple domains while maintaining reasonable complexity through the use of standardized NLP techniques

Inventive Principle:
Principle #35Parameter changes

2Adaptability or versatility

If the server processes data from multiple domains, then the system versatility improves, but the data volume increases significantly making traditional mapping approaches impractical

Engineering Contradiction:
Improvemulti-domain processing capabilityVSAvoiddata volume
Core Design Contradiction:
Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent extracts the essential meaning and intent from large volumes of multi-domain data using natural language processing. Instead of processing the entire raw data volume, the NLP system extracts key information, entities, and relationships, converting unstructured multi-domain data into structured queries that can be efficiently processed with minimal data retention

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The NLP system acts as an intermediary that reduces the complexity and volume of multi-domain data before it reaches the processing core. This intermediary layer standardizes diverse data formats from different domains into a unified query structure, making the data manageable without losing essential information

Inventive Principle:
Principle #24Intermediary (Mediator)

3Ease of operation

If pre-programmed responses are used for event processing, then the system is simple to operate, but no service is triggered when the pre-programmed response does not exist in the stored mapping

Engineering Contradiction:
Improveresponse generation capabilityVSAvoidservice trigger reliability
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The patent implements a self-service mechanism where the system automatically generates appropriate responses using natural language processing when pre-programmed responses are not available. Instead of failing when a mapping is not found, the system autonomously processes the query through NLP to generate a suitable response, ensuring continuous service availability without requiring manual intervention to add new mappings

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3398320B1System, a method and a storage medium for a unified connected network
Publication Date: 2025.08.13 SONY GROUP CORP
  • EP3398320B1 patent drawingFigure 1
  • EP3398320B1 patent drawingFigure 2A
  • EP3398320B1 patent drawingFigure 2B

AI summary

Various aspects of a system and method to provide a unified connected network are disclosed herein. The system comprises a server communicatively coupled to an electronic control unit (ECU) of a vehicle and a first set of electronic devices. Upon detection of an event, the server receives a first set of data from the ECU and/or a second set of data from the first set of electronic devices associated with a plurality of services. The server extracts information from the received first set and/or the second set of data, based on analysis of the received first set and/or the second set of data. The server further determines one or more responses that correspond to the detected event, based on the extracted information. The server further transmits the determined one or more responses to a second set of electronic devices associated with the plurality of services.