Head-Mounted Scene Data Caching for Low-Power Environment Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic devices with sensors face challenges of high power consumption and memory requirements when obtaining sensor data for their physical environment.
Innovation Solution
The device employs a strategy where it builds a scene understanding data set while stationary at a location, using motion sensors to reference this data set without activating cameras and depth sensors when the orientation remains unchanged, thereby conserving power and memory.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the device continuously activates sensors to obtain sensor data for the physical environment, then the quality and completeness of environmental data is improved, but power consumption increases
Solution Approach 1:
The device performs preliminary action by building a scene understanding data set in advance while stationary at a location. This pre-built data set contains environmental information that can be referenced later without activating sensors, thereby resolving the contradiction between data quality and power consumption.
Solution Approach 2:
The device creates a copy of the environmental scene understanding data set that can be referenced without re-capturing the actual sensor data. This copy allows the device to answer queries about the environment without activating power-consuming sensors, maintaining data quality while reducing energy usage.
2Reliability
If the device continuously activates sensors to obtain sensor data for the physical environment, then the completeness of environmental data is improved, but memory requirements increase
Solution Approach 1:
The device extracts only the essential scene understanding information from sensor data while stationary, storing a condensed data set that captures the essential environmental features. This extraction approach maintains data completeness for query purposes while reducing the memory footprint compared to storing raw continuous sensor data.
3Use of energy by moving object
If the device uses motion sensors to reference data sets without activating cameras and depth sensors, then power consumption is reduced, but the ability to capture new environmental information is limited
Solution Approach 1:
The device dynamically adapts its sensor activation strategy based on whether it is stationary or moving. When stationary, it builds scene understanding data sets; when moving within threshold distance, it uses motion sensors to reference existing data. This dynamic approach reduces power consumption while maintaining the ability to capture new environmental information when necessary.
Data Source
AI summary
A head-mounted device may include one or more sensors that obtain sensor data for a physical environment around the head-mounted device. One or more sensors such as cameras and depth sensors may be used to generate a scene understanding data set for the physical environment. The scene understanding data set may be associated with a position of the head-mounted device. While the head-mounted device is at the position associated with the scene understanding data set, the scene understanding data set may be referenced using only motion data to mitigate power consumption.