Wearable Sensor Detecting Object Emissions via Body Conduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object recognition systems face challenges in detecting and identifying objects based on their interactions, as they often require objects to be tagged with RFID tags or rely on motion-driven approaches that lack context, and are susceptible to electromagnetic noise interference.
Innovation Solution
A wearable recognition device that utilizes a sensor and controller to detect and classify electromagnetic signals emitted by objects through the human body, allowing for on-touch object detection without the need for object instrumentation, using a modified software-defined radio receiver and machine learning algorithms to differentiate between background noise and object-specific signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If RFID systems are used for object recognition, then object identification capability is improved, but all objects must be physically tagged with RFID tags which increases device complexity and cost
Solution Approach 1:
The patent converts electromagnetic noise, traditionally considered a harmful interference, into a useful signal for object recognition. By detecting and analyzing the unique electromagnetic emissions naturally produced by electrical devices during operation, the system identifies objects without requiring any tags or additional instrumentation on the objects themselves.
2Productivity
If motion-driven approaches are used for activity detection, then user activity recognition is improved, but context information about object interactions is lost
Solution Approach 1:
The patent introduces electromagnetic emissions as an intermediary signal that bridges the gap between user motion and object interaction context. The sensor detects electromagnetic signals conducted through the user's body during object interactions, providing contextual information about which objects are being touched or manipulated while maintaining the ability to recognize user activities.
3Measurement precision
If sensors are placed close to the skin for signal detection, then signal sensitivity is improved, but the sensor becomes susceptible to electromagnetic noise interference
Solution Approach 1:
The system turns electromagnetic noise from a harmful factor into the detection mechanism itself. By placing sensors close to the skin to detect electromagnetic signals conducted through the body during object interactions, the system uses the same proximity that increases noise susceptibility to also increase sensitivity to the useful electromagnetic signals from objects, then applies signal processing to distinguish between noise and meaningful signals.
4Measurement precision
If object tagging is required for recognition, then object identification accuracy is improved, but ease of operation is reduced due to calibration and setup requirements
Solution Approach 1:
The patent enables objects to serve themselves in the recognition process by naturally emitting electromagnetic signals during normal operation. The system detects these self-produced signals without requiring any tags, labels, or calibration procedures on the objects. Users simply interact with objects as normal, and the system automatically detects and identifies them through their inherent electromagnetic emissions.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables accurate and robust object recognition across various environments and users, providing context-aware information without the need for object tagging or extensive calibration, while being resistant to electromagnetic noise interference.
Implementation Method 1
The sensor may be sufficiently close to the skin of the human user to be capacitively coupled thereto
Implementation Method 2
EM noise propagates through conduction over circuits and power lines (1 kHz-30 MHz) or through radiation in free space (30 MHz to 10 GHz)
Data Source
AI summary
An object recognition device that senses electrical signals conducted by the body of a human user (e.g., as a result of direct contact with or close proximity to a device emitting or conducting electromagnetic noise), compares the sensed electrical signals to a plurality of signatures of electrical signals produced by a corresponding plurality of types of electrical and electromechanical devices to determine the type of electrical or electromechanical device that generated the electrical signals sensed by the sensor, and communicates information to the human user related to or triggered by the electrical or electromechanical device.