Position Prediction System for Geo-fencing Error Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Satellite positioning systems, such as GPS, face challenges in providing accurate location data indoors, under heavy foliage, or in urban areas, leading to erroneous position measurements that can trigger false alarms in geo-fencing applications, where users like the elderly or those with Alzheimer's may be incorrectly identified as outside or inside safe areas.
Innovation Solution
A method that determines the state of motion of a user using sensors like accelerometers and gyroscopes, predicts the device's position using multiple models, and compares it to measured positions to identify and filter out erroneous measurements, employing a Kalman filter and adjusting thresholds based on previous data to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If satellite positioning systems are used to obtain position measurements, then position data can be obtained for geo-fencing applications, but erroneous measurements occur in indoor, urban canyon, or heavy foliage environments leading to false alarms
Solution Approach 1:
The system performs preliminary actions by determining the user's state of motion (walking, running, stationary, etc.) before comparing position measurements. This preliminary classification allows the system to select appropriate prediction models and thresholds that account for the expected position changes based on motion state, thereby improving the reliability of outlier detection in challenging environments where satellite signals may be erroneous
Solution Approach 2:
The system uses feedback by comparing predicted position (based on previous position and state of motion) with actual measured position. When the measured position deviates significantly from the predicted position beyond a calculated threshold, the system identifies it as an outlier and excludes it from geo-fence evaluation. This feedback mechanism continuously refines position accuracy by eliminating erroneous measurements that cause false alarms
2Measurement precision
If multiple prediction models are used to improve position accuracy, then measurement precision improves, but device complexity increases
Solution Approach 1:
The system applies dynamics by adapting the prediction model selection and threshold values based on the real-time state of motion of the user. Different motion states (walking, running, stationary, in vehicle) have different expected position change rates, so the system dynamically adjusts which prediction models to use and what thresholds to apply. This dynamic adaptation improves position prediction accuracy without requiring a permanently complex system structure, as the complexity is activated only when needed based on motion state
Data Source
AI summary
There is provided a method of detecting erroneous measurements of the position of a device, the device comprising position measurement means, the method comprising determining the state of motion of a user of the device from a plurality of possible states; predicting the position of the device using a plurality of models and the determined state of motion of the user, each model corresponding to a respective one of the plurality of possible states; measuring the position of the device using the position measurement means; comparing the predicted position of the device to the measured position of the device; and determining whether the measured position of the device is erroneous based on the result of the step of comparing.