Edge Data Encoding for Private Low-Power Behavior Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing monitoring technologies for vulnerable individuals or environments face challenges in maintaining privacy and efficiency, as they often transmit raw data that is easily intercepted and require significant computational power, leading to inefficiencies in storage and power consumption.

Innovation Solution

Edge devices encode sensed data using autoencoders and transmit encoded representations to a computing system, which trains and redeployes updated encoders and prediction models, enhancing privacy and reducing power requirements while improving accuracy and efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If raw sensed data is transmitted to the computing system, then the computing system can process complete information for accurate behavior prediction, but the transmission consumes more power, increases storage requirements, and reduces privacy security

Engineering Contradiction:

Improvebehavior prediction accuracyVSAvoidtransmission power consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The edge device performs preliminary encoding of sensed data before transmission, transforming raw data into compressed representations. This preliminary action reduces the data volume that needs to be transmitted, thereby lowering power consumption while preserving essential information for accurate behavior prediction at the computing system

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system segments the data processing function between the edge device (encoding) and the computing system (decoding and analysis). This segmentation allows the edge device to handle only the encoding task locally, reducing transmission power requirements, while the computing system receives pre-processed data that maintains sufficient information for accurate prediction

Inventive Principle:

Principle #1Segmentation

2Loss of information

If raw sensed data is transmitted to the computing system, then complete information is available for analysis, but the data is more easily intercepted and read by unintended recipients

Engineering Contradiction:

Improveinformation completenessVSAvoidprivacy security risk

Core Design Contradiction:

Loss of informationVSObject-affected harmful factors

Solution Approach 1:

The system changes the parameter of data representation from raw sensed data to encoded data. This parameter change transforms the data into a form that is mathematically equivalent for analysis purposes but semantically obscured, making it difficult for unintended recipients to interpret even if intercepted, while preserving information completeness for authorized processing

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The encoding function acts as an intermediary between the raw sensed data and the transmitted data. This intermediary transforms the data into a protected representation that maintains information completeness for authorized systems while introducing a barrier against unauthorized interpretation, effectively mediating between information preservation and security

Inventive Principle:

Principle #24Intermediary (Mediator)

3Ease of operation

If raw sensed data is stored at the edge device, then local processing capability is maintained, but the memory constraints of edge devices are exceeded and transmission frequency must increase

Engineering Contradiction:

Improvelocal processing capabilityVSAvoiddata storage volume

Core Design Contradiction:

Ease of operationVSQuantity of substance

Solution Approach 1:

The system extracts only the essential features from raw sensed data through encoding, removing redundant information while preserving critical patterns needed for behavior prediction. This extraction reduces the storage volume required at the edge device while maintaining local processing capability for encoding operations

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS20260087312A1Combined deep learning inference and compression using sensed data

Publication Date: 2026.03.26 OPTUM INC

AI summary

An example device is configured to encode first sensed data using a first encoder and to predict a first behavior based on the encoded first sensed data to create a first prediction using a first prediction model. The example device is configured to store the encoded first sensed data in the one or more memory units. The example device is configured to control the communication unit to transmit the encoded first sensed data in a first batch to a computing system. The example device is configured to receive, from the computing system via the communication unit, a second encoder, the second encoder being based at least in part on the encoded first sensed data. The example device is also configured to receive, from the computing system via the communication unit, a second prediction model, the second prediction model being based at least in part on the encoded first sensed data.