Population-Based Energy Footprint Reconfiguration for Apps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to assess and identify device configuration alterations that can lead to energy savings due to complex energy-consumption dependencies between applications and underlying hardware and software configurations, making it difficult to compare or assess the energy efficiency of different applications or their versions across devices.
Innovation Solution
A method is disclosed that determines a user device energy footprint based on energy consumption data, collects data from a population of devices with similar configurations, and generates reconfiguration recommendations to match the configurations with lower energy footprints, using a population-based energy savings discovery engine to identify potential energy savings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If device-level power efficiency assessment is performed based on measured power consumption, then energy efficiency can be evaluated for individual devices, but it becomes difficult to compare energy efficiency across different applications or versions due to hardware and software configuration variations
Solution Approach 1:
The patent introduces software as an intermediary layer between hardware and applications. By measuring and comparing power consumption at the software level (OS and drivers) rather than directly at hardware level, the system can isolate software-related energy consumption from hardware-specific variations, enabling fair comparisons across different devices running the same application
Solution Approach 2:
The patent creates a standardized measurement model that copies the same measurement methodology across different devices. By using consistent measurement procedures, data collection methods, and analysis approaches, the system enables comparable energy efficiency assessments across diverse hardware configurations without being constrained by device-specific differences
2Use of energy by moving object
If total energy consumption of an application is measured on a user device, then energy footprint can be determined, but the measurement becomes tightly intertwined with hardware and software configuration making it difficult to identify energy-saving configurations
Solution Approach 1:
The patent segments the total energy consumption into distinct components: application energy consumption, operating system energy consumption, and driver energy consumption. By dividing the measurement into these separable parts, the system can identify which component contributes most to energy usage and target optimization efforts accordingly, rather than treating total consumption as an inseparable whole
Solution Approach 2:
The patent implements feedback mechanisms where energy consumption data from multiple devices and configurations is collected, analyzed, and used to generate recommendations. The system continuously gathers measurement data, compares it across different configurations, and feeds back optimized configuration recommendations to users, creating a closed-loop system for energy efficiency improvement
Data Source
AI summary
A disclosed method facilitates identification and recommendation of alternative device configuration(s) with the potential to reduce or mitigate the energy footprint of a particular application on a user device. A disclosed method includes determining a user device energy footprint for an application executing on a user device and determining a per-device energy footprint for the application with respect to a population of user devices executing the application that are characterized by a device configuration not shared by the user device. An energy savings metric is determined based on a comparison between the per-device energy footprint within the population and the user device energy footprint. In response to determining that the energy savings metric is indicative of a potential energy savings, an energy savings reconfiguration recommendation is generated.