Mobile Device Posture Correction via Neural Network Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for correcting body posture and recommending exercises to alleviate physical stress from using mobile devices lack real-time, hassle-free analysis and often require multiple wearables or rely on an inactive camera for continuous monitoring.
Innovation Solution
A method and system that utilize inertial sensor data and touch screen data from a user equipment (UE) to determine the holding orientation, body posture, and impacted body parts, predicting the impact level and recommending exercises for posture correction through a neural network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple wearables are used to capture body posture, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent combines multiple sensing capabilities (accelerometer, gyroscope, magnetometer, camera, touch screen) into a single integrated system within the mobile device. This merging of functions allows the device to capture body posture through multiple sensors simultaneously, achieving high measurement precision without requiring multiple separate wearables.
Solution Approach 2:
The mobile device is designed to perform multiple functions: it serves as both a computing device and a posture monitoring system. The inertial sensors, camera, and touch screen work together to provide comprehensive body posture analysis, making the device universal rather than requiring specialized wearables for each function.
2Measurement precision
If camera is used to capture body posture, then measurement precision is improved, but ease of operation deteriorates due to camera inactivity
Solution Approach 1:
The system activates the camera in advance based on predictions from inertial sensors. When the sensors detect a posture that warrants monitoring, the camera is triggered to capture images before the user becomes aware of the posture issue, ensuring continuous monitoring capability without requiring the camera to remain constantly active.
Solution Approach 2:
The system uses feedback from inertial sensors to control camera activation. The sensors continuously monitor device orientation and trigger the camera when specific posture conditions are detected, creating a feedback loop that ensures the camera operates only when needed for posture correction.
3Productivity
If real-time posture analysis is implemented, then productivity is improved, but use of energy increases
Solution Approach 1:
The system implements periodic posture analysis rather than continuous monitoring. The inertial sensors operate continuously at low power to detect posture changes, and the energy-intensive camera is activated periodically only when posture correction is needed, reducing overall energy consumption while maintaining real-time monitoring capability.
Solution Approach 2:
The system uses partial action by activating the camera only for specific posture scenarios rather than continuously. The inertial sensors provide continuous monitoring at low power, and the camera is engaged partially only when posture issues are detected, optimizing the balance between real-time analysis and energy consumption.
Data Source
AI summary
A method may include: receiving inertial sensor data and touch screen data of the user equipment (UE); determining an application type running on the UE; predicting, by a neural network, a holding orientation of the UE based on the inertial sensor data, the application type, and the touch screen data; determining, by the neural network, a body posture of the user and at least one impacted body part based on the inertial sensor data, based on the holding orientation; determining, by the neural network, an impact level of the at least one impacted body part based on the body posture, the holding orientation of the UE and the inertial sensor data of the UE; and recommending a body posture correction and the at least one exercise for the at least one impacted body part based on the impact level.