Mobile device and system for automated trip familiarity recognition and corresponding method thereof

1 . A method for electronic trip familiarity detection based on sensory data measured by a plurality of sensors of a mobile device associated with a vehicle, the sensors comprising a GPS sensor and an accelerometer, the mobile device acting as a wireless node within a cellular data transmission network by means of antenna connections of the mobile device to the cellular data transmission network, the GPS and accelerometer being connected to a monitoring mobile node application of the mobile device capturing usage-based and/or user-based sensory data of the GPS and accelerometer of the mobile device, the method comprising:

measuring a plurality of time series of sensory parameter values based on measuring parameters obtained from the GPS and accelerometer of the mobile device, each of the measured time series at least comprising a time series of the sensory parameter values including location and speed measurements of the GPS sensor, and the GPS sensor measuring longitude, latitude, and altitude positions of the mobile device as the time series by measuring different speeds of light delays in signals received from two or more satellites;

detecting, for each of the measured time series, a start point and an end point of at least a part of the measured time series of the sensory parameter values, the detection of the start point and the end point triggering allocation of the at least part of the measured series to a trip having the detected start point and the detected end point, the detection of the start point and the end point in the at least part of the measured time series of the sensory parameter values including reporting exact geo-coordinates by a monitoring module when a start recording trigger is fired and when a stop recording trigger is fired, the start recording trigger and the stop recording trigger being fired based on an edge module running in the mobile device monitoring different sensory parameters including at least parameters from motion activity detection, the start recording trigger being triggered by at least detecting a significant location change, and the stop recording trigger being triggered by at least detecting a GPS speed below a threshold value for a predefined time period;

in response to the start recording trigger being fired, processing a trip validation by monitoring GPS speed during a time period below or equal to 120 seconds, wherein

when there are at least 5 GPS speed values above or equal to 20 km/h within 120 seconds, the trip validation process is successfully finished and a trip recording stage starts, and

during the trip recording stage, the mobile node application records GPS sensory data at 1 Hz and accelerometer sensory data at 10 Hz, the recording being stopped when the stop recording trigger is fired;

measuring, for any pairing of trips, a first geographical distance between the detected start points of said trip pairing and between the end points of said trip pairing based on location measurement parameter values of the sensory parameters of the at least part of the measured time series and a second geographical distance between the detected start point of a first trip and the detected endpoint of a second trip of said trip pairing and between the detected start point of the second trip and the detected endpoint of the first trip of said trip pairing based on the location measurement parameter values of the sensory parameters of the at least part of the measured time series;

measuring a third geographical distance by measuring, for each of the trips of the pairing a trip length between the detected start and end points of each trip of the trip pairing, the third geographical distance being given by a relative difference of the measured trip lengths;

mutually normalizing the first, second, and third geographical distances by applying a modified sigmoid transfer mapping to the first, second, and third geographical distances in a normalized range from 0 and 1; and

measuring an overall familiarity parameter value by composing the first, second, and third normalized geographical distance values to a weighted average of the first, second, and third normalized geographical distance values,

wherein an overall familiarity parameter value of 0 indicates identical trips and an overall familiarity parameter value of 1 indicates completely different trips.

2 . The method for electronic trip familiarity detection according to claim 1 , wherein clusters of trips are stored having start and end regions with a defined radius.

3 . The method for electronic trip familiarity detection according to claim 2 , wherein a center of a start and/or end region is defined by an average of all start and end points of trips associated with a cluster and the radius is defined as a respective standard deviation.

4 . The method for electronic trip familiarity detection according to claim 2 , wherein, for a cluster, an average travelled distance and its standard deviation are stored.

5 . The method for electronic trip familiarity detection according to claim 4 , wherein a trip is matched by generating a distance between start and/or end points in units of the respective radius and a distance in terms of travelled kilometers in units of the standard deviation.

6 . The method for electronic trip familiarity detection according to claim 2 , wherein the trips are clustered using an agglomerative clustering by a machine learning structure grouping iteratively together trips that are closest together until there are no trips left that are closer together than a defined threshold.

7 . The method for electronic trip familiarity detection according to claim 6 , further comprising using one or more predefined hyperparameter dynamically optimized in dependence of increasing and decreasing performance measures.

8 . The method for electronic trip familiarity detection according to claim 7 , wherein, for generation of the first, second, and/or third geographic distance, the method includes using at least two hyperparameters, which are weights for the start and end point and for the travelled distance.

9 . The method for electronic trip familiarity detection according to claim 7 , wherein, for the clustering, the method includes using at least two hyperparameters, which are:

a distance threshold, wherein within its range two trips are detected to form a cluster, and

a minimum number threshold of trips required to create a new cluster.

10 . The method for electronic trip familiarity detection according to claim 7 , wherein the optimization of the one or more hyperparameter is performed separately for each possible technical application comprising different modes of transportation and/or different geographic locations and/or driver to passenger trips.

11 . The method for electronic trip familiarity detection according to claim 1 , wherein the mobile device is a mobile telecommunication device GPS and accelerometer are integrated sensors of said mobile telecommunication device.

12 . The method for electronic trip familiarity detection according to claim 11 , wherein the mobile telecommunication device is a cellular mobile phone and/or a smart phone.

13 . The method for electronic trip familiarity detection according to claim 1 , wherein transport modes distinguishable by the method include at least land-based vehicles and/or maritime vehicles and/or air-based transportation means and/or hiking and/or skiing.

14 . The method for electronic trip familiarity detection according to claim 13 , wherein the land-based vehicles at least comprise distinguishable modes for cars and/or motor bikes and/or bicycles.

15 . The method for electronic trip familiarity detection according to claim 1 , wherein the detection of the start point and the end point in the at least part of the measured time series of the sensory parameter values includes selecting earliest geo coordinates for the start point and latest geo coordinates for the end point.

16 . The method for electronic trip familiarity detection according to claim 1 , wherein the detection of the start point and the end point in the at least part of the measured time series of the sensory parameter values includes reporting of the exact geo coordinates by an application when the start recording trigger is fired and when the stop recording trigger is fired.