Activity Recognition via Adaptive Sensor Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electronic devices struggle to accurately recognize irregular and high-intensity user activities, such as sports, due to their reliance on regular movement patterns detected by acceleration sensors, which leads to misclassification of exercises and inefficient power consumption.
Innovation Solution
The method employs a first processor to continuously monitor sensor data from multiple sensors, determining user activity based on the intensity and irregularity of movement patterns, activating additional sensors only when regular patterns are detected to conserve power and improve accuracy in identifying exercises like basketball or soccer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If acceleration sensor data is used to detect regular movement patterns, then power consumption is reduced, but irregular exercises cannot be recognized
Solution Approach 1:
The system dynamically adjusts the detection strategy based on the characteristics of movement patterns. For regular patterns, it uses acceleration sensor data only. For irregular patterns, it transitions to using heart rate sensor data, making the detection system adaptive rather than static.
Solution Approach 2:
The system changes the detection parameters (from acceleration-based to heart rate-based) depending on the type of activity detected. This parameter switching allows the system to maintain low power consumption for common activities while accurately detecting irregular exercises when needed.
2Reliability
If heart rate sensor is continuously activated to detect all exercise types, then exercise recognition accuracy is improved, but power consumption increases
Solution Approach 1:
Instead of continuous activation, the heart rate sensor is activated periodically or conditionally only when irregular exercise patterns are detected by the acceleration sensor, reducing overall power consumption while maintaining detection accuracy.
Solution Approach 2:
The acceleration sensor performs the initial screening and self-identifies when its capability is insufficient (irregular patterns), automatically triggering the heart rate sensor to supplement the detection, creating a self-managing multi-sensor system.
3Use of energy by moving object
If only regular movement patterns are detected, then power consumption is minimized, but irregular exercises are misclassified
Solution Approach 1:
The detection system is segmented into two independent but complementary pathways: one for regular patterns using acceleration data, and another for irregular patterns using heart rate data. This segmentation allows each pathway to be optimized for its specific function.
Solution Approach 2:
The acceleration sensor acts as an intermediary that first analyzes the movement pattern and determines whether to activate the heart rate sensor based on the detected pattern's regularity, serving as a gatekeeper that prevents unnecessary heart rate measurements.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Various exemplary embodiments provide an electronic device configured to include one or more sensor modules, at least one memory, a display, and a first processor or second processor operatively coupled to the one or more sensor modules, the at least one memory, and/or the display, wherein the first processor is configured to acquire sensor data from at least one sensor module among the one or more sensor modules, calculate a difference between at least two data points in the acquired sensor data, determine user's activity information based on at least one of a period of the sensor data, a magnitude of the difference, and a change amount of the difference, and deliver the user's activity information to the second processor, and wherein the second processor is configured to display, on the display, a user interface related to the user's activity information delivered from the first processor. Other exemplary embodiments are also possible.