Assistant Client Caching for Low-Latency Smart Device Control

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing automated assistant systems face high latency and excessive resource usage due to the need for remote processing of user inputs to control smart devices, which is exacerbated by geographical distance and bandwidth requirements.

Innovation Solution

Implementing a caching mechanism on client devices that stores locally interpretable semantic representations of user inputs, allowing for local control of smart devices without the need for constant remote processing, and using cloud semantic representations only when necessary.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If remote automated assistant servers process user inputs to control smart devices, then comprehensive control capability is achieved, but latency increases and network resources are excessively consumed

Engineering Contradiction:
Improvecontrol capabilityVSAvoidlatency
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The patent segments the automated assistant system into local and remote components. The local automated assistant client handles routine smart device control tasks independently, while the remote server handles complex tasks requiring cloud semantic representations. This segmentation allows common operations to execute locally without network latency, while maintaining comprehensive control capability through cloud backup.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements preliminary action by storing cloud semantic representations locally in a cache before they are needed. When a user input matches a cached representation, the system can immediately execute the corresponding smart device control without waiting for remote server processing. This pre-caching mechanism eliminates latency for repeated or predictable commands.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If remote servers process all user inputs, then accurate semantic understanding is achieved, but server resources and network bandwidth are excessively used

Engineering Contradiction:
Improvesemantic understanding accuracyVSAvoidnetwork resource consumption
Core Design Contradiction:
Measurement precisionVSLoss of energy

Solution Approach 1:

The patent applies partial action by having the local automated assistant client handle only those user inputs that match locally cached cloud semantic representations. Remote server processing is invoked only when local caching fails to provide a match. This selective approach maintains accurate semantic understanding for common patterns while minimizing network resource consumption.

Inventive Principle:
Principle #16Partial or excessive action

3Productivity

If local processing is implemented on client devices, then latency is reduced and network resources are conserved, but processing capability is limited by device resources

Engineering Contradiction:
Improveprocessing speedVSAvoidprocessing capability
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent introduces cloud semantic representations as intermediaries between user inputs and smart device control. These pre-processed semantic representations are cached locally and serve as a bridge, enabling fast local execution without requiring full remote server processing capability. The intermediary contains the complexity of semantic understanding while keeping the local client device requirements minimal.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3834199B1Efficient and low latency automated assistant control of smart devices
Publication Date: 2023.04.05 GOOGLE LLC
  • EP3834199B1 patent drawingFigure 1
  • EP3834199B1 patent drawingFigure 2
  • EP3834199B1 patent drawingFigure 3A

AI summary

Various implementations relate to techniques, for controlling smart devices, that are low latency and/or that provide computational efficiencies (client and/or server) and/or network efficiencies. Those implementations relate to generating and/or utilizing cache entries, of a cache that is stored locally at an assistant client device, in control of various smart devices (e.g., smart lights, smart thermostats, smart plugs, smart appliances, smart routers, etc.). Each of the cache entries includes a mapping of text to one or more corresponding semantic representations.