Radar Data Pre-Processing for Low-False-Alarm Presence Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current smart doorbell solutions rely heavily on camera-based motion detection, leading to high false positives and negatives, increased power consumption, and reduced battery life due to unreliable human presence detection, which affects user experience and market adoption.
Innovation Solution
A radar and spiking neural network system for pre-processing radar data to detect human presence, allowing for low-power, motion-resistant detection with a low rate of false alarms, by using a 60 GHz Continuous Wave Frequency Modulated radar and a Spiking Neural Processor to analyze spatio-temporal data patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If camera-based motion detection is used for human presence detection, then the device can monitor the entrance area, but false positives increase due to lighting changes, wind-triggered movements, rain, dust, and insects
Solution Approach 1:
The patent segments the detection task across multiple independent sensors (camera, PiR sensor, radar) rather than relying on a single camera-based motion detection system. Each sensor type detects different physical phenomena (visual motion, thermal radiation, radio wave reflection), allowing the system to cross-validate detections and filter out false positives from any single sensor modality.
Solution Approach 2:
The patent introduces passive infrared (PiR) sensors and radar as intermediary detection layers between the camera and the final detection decision. These intermediaries provide additional verification - PiR sensors detect thermal signatures of living beings while radar detects motion patterns - thereby mediating the detection process to eliminate false alarms from lighting changes, dust, or insects that might trigger the camera alone.
2Reliability
If multiple sensors (PiR and radar) are used to mitigate false positives, then detection reliability improves, but the Bill of Materials and power consumption increase
Solution Approach 1:
The patent implements periodic sensing and processing cycles where the radar and PiR sensors operate in alternating or coordinated intervals rather than continuously. The system performs detection in phases - using low-power sensors periodically to screen for presence, then activating higher-power components only when needed - thereby reducing overall power consumption while maintaining reliable detection across multiple sensor types.
Solution Approach 2:
The patent enables the multi-sensor system to self-regulate power consumption by having sensors operate autonomously in low-power modes and only activating full processing chains when detection events are suspected. The system serves itself by intelligently managing when each sensor type operates at full power versus standby, eliminating the need for continuous high-power operation of all sensors.
3Productivity
If movement-based triggering is used, then the device can detect activity, but detection reliability decreases due to increased false positives from non-human movement
Solution Approach 1:
The patent applies dynamic analysis to motion detection by examining the temporal and spatial characteristics of detected movements. Instead of simple motion triggering, the system analyzes movement patterns, velocity profiles, and trajectory consistency over time to distinguish human motion from environmental disturbances. This dynamic approach allows responsive detection while filtering out false positives from wind-blown objects or animal movement.
Solution Approach 2:
The patent changes the detection parameters from simple motion presence to multi-dimensional motion characteristics including velocity, acceleration, trajectory, and temporal consistency. By adjusting these parameters dynamically based on context and cross-referencing multiple sensor inputs, the system maintains high responsiveness to genuine human presence while rejecting false positives that don't match human movement patterns.
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
The system achieves robust human presence detection with high accuracy (>99%) and significantly extends battery life by operating at ultra-low power, reducing false alarms and maintaining detection latency without compromising design aesthetics or buildability.
Implementation Method 1
Radar uses radio waves to determine the distance (ranging), angle (azimuth and elevation), and radial velocity of objects relative to the radar system site
Implementation Method 2
Radio waves (pulsed or continuous) from the transmitter reflect off the objects and return to the receiver, giving information about the objects' locations and speeds
Implementation Method 3
60 GHz Continuous Wave Frequency Modulated radar
Data Source
Figure 1~2
Figure 3~4
AI summary
The present invention discloses a method for pre-processing radar data for further processing by a machine learning model, the method comprising: acquiring radar data from a transmitted signal and a received signal of a radar, wherein the radar data comprises frames and/or chirps; allocating the radar data to a plurality of frequency range bins according to the radar frequency of radar data of the received signal; calculating a difference between the radar data at two or more different time steps for each of the plurality of frequency range bins, wherein the time steps correspond to the frames or chirps of the radar data; and providing the calculated differences to the machine learning model with temporal dynamics for further processing.