Method and system for the characterization of transit

Abstract

A method for the characterization of transit, comprising: obtaining, over time, report signals provided by a plurality of mobile devices within an area of interest, each of said report signals comprising a timestamp, a location parameter, and at least one of a temporary identifier and a session identifier; defining a cluster of at least a subset of said plurality of mobile devices based on the timestamp, the location parameter, and the at least one of a temporary identifier and a session identifier included in said report signals; determining, based on said report signals obtained over time, a tract followed by the cluster; characterizing a state of said cluster based on said report signals obtained over time.

Description

[0001] "METHOD AND SYSTEM FOR THE CHARACTERIZATION OF TRANSIT" DESCRIPTION

[0002] Background of the invention

[0003] Field of the invention

[0004] The present invention relates to a method for the characterization of transit.

[0005] Additionally, the invention refers to a system for the characterization of transit.

[0006] Description of the related art

[0007] In today's Zeitgeist, the usage and adoption of public transit is becoming increasingly more central in governmental policy. To improve and maximize the potential of a transit system, it is important to understand and track the usage and state of individual vehicles. Utilizing mobile device data to cluster users, therefore, may provide a benefit to transit operators in understanding the usage and state of a vehicle.

[0008] Article "Accurate, Low-Energy Trajectory Mapping for Mobile Devices" by Thiagaranjan et al., discloses a method for performing user tracking based on signal strength of a mobile network along with additional refinement processing involving auxiliary sensors incorporated on the user device.

[0009] The Applicant notes as well that Document WO 2006 / 083535 discloses a methodology for clustering mobile devices in a wireless network.

[0010] A major hurdle to obtaining the mobile device data needed for clustering is the consent request by a service provider to aggregate the necessary data. The Applicant notes that tracking requires user authorization or consent to collect and process this information. This information, therefore, is not always available and is cumbersome for the users of the device to define and / or revoke consent for individual services.

[0011] In view of the related art then, the Applicant has tackled the problem of classifying users as members of a cluster based on their movement, in a precise and accurate manner, without violating user privacy or requesting any additional authorization.

[0012] Summary of the invention

[0013] The Applicant has recognized that by employing temporary identifiers (e.g. the Temporary Mobile Subscriber Identity, TMSI), already made available by telecommunications protocols, movement can be tracked, and several types of analysis can be performed from a statistical point of view, fully respecting users' data privacy and avoiding, at the same time, any kind of authorizations from users. Moreover, TMSI data could be further anonymized.

[0014] Accordingly, in a first aspect, the present invention relates to a method for the characterization of transit.

[0015] According to an embodiment of the present invention, the method comprises determining an area of interest.

[0016] According to an embodiment of the present invention, the method comprises determining a time period of interest.

[0017] According to an embodiment of the present invention, the method comprises obtaining, over time, report signals provided by a plurality of mobile devices within the area of interest, in the time period of interest.

[0018] According to an embodiment of the present invention, each of said report signals provided by a corresponding mobile device is comprised of an identifier, a timestamp, and a location.

[0019] According to an embodiment of the present invention, the method comprises defining a cluster of at least a subset of said plurality of mobile devices based on the identifier, the timestamp, and the location parameter included in the report signals.

[0020] According to an embodiment of the present invention, the method comprises determining, based on said report signals obtained over time, a tract followed by the cluster.

[0021] According to an embodiment of the present invention, the method comprises characterizing a state of said cluster based on said report signals obtained over time.

[0022] In a second aspect, the present invention relates to a system for the characterization of transit.

[0023] According to an embodiment of the present invention, the system comprises a data processing unit configured to perform the aforesaid method.

[0024] In one or more of the above aspects, the invention can comprise one or more of the following preferred features.

[0025] According to an embodiment of the present invention, said report signals further comprise signal characteristics.

[0026] According to an embodiment of the present invention, said signal characteristics comprise one or more of: a current service cell ID, a current service cell received power value, a current service cell received quality value, and a plurality of adjacent cell IDs, a plurality of adjacent cell received power values, and a plurality of adjacent cell received quality values.

[0027] According to an embodiment of the present invention, said cluster is defined based on a minimum persistence time of mobile devices in the cluster.

[0028] According to an embodiment of the present invention, said cluster is defined based on a minimum cluster dimension.

[0029] According to an embodiment of the present invention, said cluster is defined based on an entry / exit tolerance.

[0030] According to an embodiment of the present invention, the method comprises determining a turnover of said cluster.

[0031] According to an embodiment of the present invention, said turnover is defined by a joining action and / or an exit action performed by one or more of said mobile devices.

[0032] According to an embodiment of the present invention, the method comprises determining, based at least in part upon the turnover, a set of time-based statistics for said state of said cluster.

[0033] According to an embodiment of the present invention, said set of time-based statistics comprises an average number of mobile devices in said cluster.

[0034] According to an embodiment of the present invention, said average number of mobile devices depicts said state of the cluster in a time period.

[0035] According to an embodiment of the present invention, the report signals are sent from said mobile devices according to the Minimization of Drive Test, MDT, technology.

[0036] According to an embodiment of the present invention, said tract is at least a part of an overall path of a vehicle.

[0037] According to an embodiment of the present invention, the method characterizes the state of the vehicle based on the state of said cluster.

[0038] According to an embodiment of the present invention, one of said plurality of mobile devices belongs to said vehicle defining a reference tract, said reference tract providing a reference point for defining said cluster.

[0039] According to an embodiment of the present invention, the method comprises, based on said state of the vehicle, determining a set of timebased statistics describing a level of congestion aboard said vehicle.

[0040] According to an embodiment of the present invention, the method comprises furnishing said set of time-based statistics to an end user to describe a corresponding current state and / or a historical state of said vehicle.

[0041] Brief description of the drawings

[0042] Further features and advantages will appear more clearly from the detailed description of preferred and non-exclusive embodiments of the invention. This description is provided hereinafter with reference to the accompanying illustrative and non-limiting figures, in which:

[0043] Figure 1 exemplifies the clustering process on board of a transit vehicle;

[0044] Figure 2 schematically shows a flow chart describing an exemplary embodiment of the present invention.

[0045] Description of embodiments of the present invention

[0046] The present invention is directed to a method and a system designed to characterize transit using mobile device signals.

[0047] With reference to figure 1, the system 100 is an exemplary embodiment of a system capable of performing the transit characterization.

[0048] The system 100 operates on an area of interest 1000.

[0049] The area of interest 1000, for example, can be delimited by a polygonal chain. The polygonal chain can be a simple polygonal chain or a closed polygonal chain. Preferably the polygonal chain does not include selfintersections. In an embodiment, the area of interest is defined by a polygonal chain combined with one or more confidence bands, preferably two confidence bands located at opposite sides of the same polygonal chain. The latter solution can be useful, for example, in case the area of interest is to cover a tract of a street: the polygonal chain is drawn on the street, and the two confidence bands are defined along the street, before and after the polygonal chain; in this way, the area of interest 1000 will be focused on the target of real interest (i.e. transit of people along the street, as will be clearer in the following) and will exclude noise possibly generated by mobile devices located on the sidewalks and / or in buildings on the sides of the street.

[0050] An area of interest 1000 includes a plurality of individual mobile devices 103.

[0051] The mobile devices 103 can be, for example, smartphones and / or tablets.

[0052] Each mobile device 103 is provided with all the necessary hardware and software resources in order to connect a telecommunications network 1010, specifically a Wide Area Network, WAN. Each mobile device 103 is provided with a localization module, which enables the mobile device to determine its position based on a satellite system.

[0053] Each mobile device 103 receives location signals from a Global Navigation Satellite System, GNSS.

[0054] Although specific reference is made to GNSS (GPS, Galileo, GLONASS, BEIDOU, etc.), in cases where the global positioning system is unavailable or unreliable, other methodologies may be optionally employed to determine a position of the mobile device including position detection techniques based upon cellular network data or inertial navigational methods or the like.

[0055] The mobile devices 103 combine the location signals received from the GNSS with additional information comprising a timestamp and at least one of a temporary identifier and a session identifier. This information is then provided to the telecommunications network 1010 by means of report signals 106.

[0056] The telecommunications network 1010 may, e.g., comprise a cellular network utilizing 4G and / or 5G technology. The telecommunication network may optionally comprise other cellular network technologies.

[0057] The telecommunication network 1010 comprises, for example, base stations 110, 111, 112.

[0058] Base stations 110, 111, 112 receive the report signals 106 and provide the same to further components of the network - which will be disclosed in the following.

[0059] Timestamp data included in the report signals 106 indicates the date (preferred format: dd-mm-yyyy) and time (preferred format: hh:mm:ss.cc) in which the mobile device 103 detected / calculated its own geographic position.

[0060] The temporary identifier is an identifier which is randomly assigned to a mobile device when it connects to a mobile network. In a 4G network, for example, the temporary identifier is referred to as TMSI - Temporary Mobile Subscriber Identity. The temporary identifier replaces the IMSI - International Mobile Subscriber Identity - when communicating with the network to hide or minimize the chances of leaking the IMSI to bad actors. As known in the art, the IMSI uniquely identifies a subscription to a mobile network and is rarely used, to protect users and their privacy. Therefore, the IMSI is replaced by the TMSI which is regenerated at each new connection to the network, including when exiting or re-entering the network, or after each restart of the mobile device 103. The TMSI remains allocated to the same mobile device 103 without changing as long as the associated mobile device remains connected to the access network, even in cases of handover, until a change of Location Area or Tracking Area occurs.

[0061] Optionally, in cases where user data cannot be anonymized through the usage of the TMSI, the method may instead use a non-temporary identifier. This non-temporary identifier may be a IMSI or another similar persistent identifier. In this case, additional steps are required to address privacy concerns of users including requesting user consent for the processing of this data in accordance with the applicable state / national laws for the area of interest. Advantageously, the usage of a non-temporary identifier like the IMSI allows for higher accuracy over longer time periods as the number of data points with which to perform the method increases due to the persistence of the identifying parameter. On the contrary, when the temporary identifier / session identifier is regenerated, the signals can no longer be associated with a specific mobile device. This improvement in signal continuity can be shown in the case of a change of radio access technology. If, for example, a mobile device connected to a 4G network during a first part of a tract switches temporarily to a 5G network for a second part of the tract, then back to the 4G network for a third part of the tract, the TMSI would need to be regenerated and the data for the entire tract will not be available due to the loss of continuity that occurred. If IMSI identifying signals are used, this loss of continuity will not occur.

[0062] The session identifier identifies a session (e.g. a call, an SMS exchange, application data exchange, etc.) generated by a determined mobile device, based on a determined subscription, in a determined time period. Preferably, the time period is defined by the MNO (Mobile Network Operator) according to the legislation in force. During a session, the mobile device typically sends multiple report signals, each including the same session identifier. The frequency at which the mobile device generates the report signals is preferably defined by the MNO, determining a reasonable trade-off between accuracy (increases with number of report signals) and traffic overhead (can be limited by decreasing the number of report signals).

[0063] Preferably, each mobile device 103 generates useful information while in Idle mode. The information remains on the device locally until such time as it can be transmitted to the network through the report signals 106 when the mobile device switches to a connected state with the network.

[0064] In a particularly advantageous embodiment, the report signals 106 are MDT (Minimization of Drive Test) signals. Such technology is standardized by ETSI and disclosed in technical specification 3GPP TS 37.320; for example, version 15.0.0 (2018-06) can be taken into consideration. As an example, the MDT information sent to the network may include the following:

[0065] The report signals 106 are generated in various timesteps, represented in figure 1 by Tl, T2, and T3, and provided to the network 1010.

[0066] According to the present invention, the report signals 106 considered for further processing are obtained in a time period of interest.

[0067] The time period of interest is preferably defined between a start time and an end time.

[0068] Thus, in view of the above, the processing that will be disclosed in the following is based on report signals 106 generated, in the time period of interest, by mobile devices 103 which are located in the area of interest 1000.

[0069] In the exemplary representation of figure 1, timesteps Tl, T2, T3 are included in the time period of interest.

[0070] In the exemplary representation of figure 1, report signals 106 at time Tl are received by base station 110, report signals 106 at time T2 are received by base station 111 and report signals at time T3 are received by base station 112. However, it has to be noted this is just an example, and a given base station can receive report signals 106 at different times.

[0071] The information in time provided to the network 1010 is then passed on to the data processing unit 1020. The data processing unit 1020 is configured to receive the report signals 106 and then perform further analysis upon the signals. The data processing unit 1020 may be further subdivided into a central processing unit 121 and an associated storage system 120.

[0072] Preferably, the further analysis performed by the data processing unit 1020 includes a preliminary correlation operation. The Applicant observes that the temporary identifier (e.g. the TMSI) and the session identifier are not both always included in each report signal provided to the network. The same temporary identifier (e.g. TMSI) is used over a longer period than the session identifier but is transmitted only at the beginning of a connection, whereas the session identifier is updated / changed at each user session. Therefore, to maintain the longest time frame associated with each mobile device, an operation is performed to associate / correlate the report signals with each other, based on the temporary identifier and / or session identifier included therein. Thusly, sequences of correlated data (e.g. MDT samples) may be created.

[0073] The report signals provided, in time, by each mobile device 103, allow for a tract to be created based on the location of the mobile device in time. Through well-known manipulations, velocity, acceleration, and other relevant data points may be obtained over the tract as well.

[0074] A tract travelled by a given mobile device is defined by the sequence of positions, in time, of such mobile device.

[0075] The data processing unit 1020, based upon the tracts defined for each of the mobile devices 103, performs a clustering operation. The clustering operation may be performed through any clustering / classification methodology based on the similarities and differences between the timevarying signals obtained and their respective tracts. In the case of vehicle, the Applicant notes that the individuals inside the vehicle will all have very similar trajectory, speed, acceleration, and / or other characteristics over a common tract.

[0076] The Applicant observes that, in the present context and the appended claims, the term "tract", when referring to a cluster, indicates a path or path portion travelled by the cluster.

[0077] The clustering operation may include the use of additional parameters.

[0078] The additional parameters are advantageous in order to more reliably and accurately define the cluster and / or to minimize the computational complexity.

[0079] The additional parameters may include one or more of: a minimum cluster dimension, a minimum persistence time, and / or an entry / exit tolerance.

[0080] A minimum cluster dimension parameter defines a minimum number of mobile devices to consider as a potential cluster for the clustering operation.

[0081] A minimum persistence time in is defined as a parameter defining a minimum time length in which the cluster must be valid in time. A valid cluster is a cluster which includes a minimum number of mobile devices, i.e. a cluster which has a minimum cluster dimension. This parameter eliminates short-lived clusters that may otherwise cloud the subsequent analysis, and lets such analysis focus on clusters which survive for longer time periods and are therefore more likely to be significant in a subsequent processing.

[0082] An entry / exit tolerance is defined as the number of mobile devices able to enter into or exit from a previously defined cluster before the cluster must be redefined. In other words, a cluster is considered to be still itself (and not a new, different cluster) if the number of mobile devices that enter into I exit from the cluster remains below a preset threshold, which represents the above-mentioned entry / exit tolerance. In an embodiment, the entry / exit tolerance is defined as a number of mobile devices. In an embodiment, the entry / exit tolerance is defined as a number of mobile devices over a given time interval.

[0083] These additional parameters may be used singularly or in combination with other parameters to improve accuracy of the system and / or to lessen the complexity and computation cost of the significant number of mobile devices and signals in a given area and time period of interest.

[0084] In cases of heavy traffic for example, situations may occur where variations in the location parameters provided by numerous mobile devices in congested conditions are mistaken or add noise to a valid cluster of interest. To avoid this problem, additional cluster parameters such as those previously mentioned may become advantageous for the processing of the data.

[0085] Optionally, the mobile device characteristics used for clustering may include a network service cell and / or a number of additional adjacent cells and the associated network parameters. These network parameters may include Reference Signals Received Power (RSRP) and Reference Signals Received Quality (RSRQ) among other network parameters.

[0086] The Applicant notes that in some cases, some vehicles may be already equipped with a mobile device. In this case, the report signals from the mobile device associated with the vehicle may be used as a reference tract. The tract followed in time by the mobile devices associated with the vehicle can be used to refine / define the clustering operation and provide the base signal properties to compare with the signals from the mobile devices.

[0087] The cluster 101 resulting from the clustering operation performed by the data processing unit 1020 allows for different sets of mobile devices to be defined, i.e. the mobile devices belonging to the cluster 101, and the mobile devices nearby but not members of the cluster 101.

[0088] Preferably, the clustering operation allows for the definition of a turnover of the cluster. Based on the similarity between tracts and the timebased membership in the cluster, joining actions and exit actions by users carrying the mobile devices can be defined. For example, mobile device 105, shown in figure 1, was part of the cluster 101 at times T1-T2 and at time T3 has left the cluster 101; processing of subsequent report signals 106 provided by the mobile devices still included in the cluster and by the mobile device 105 that has left, will reveal that mobile device 105 shows positional / movement features which are not compatible with the movement of the cluster. Accordingly, the data processing unit 1020 will come to the conclusion that the cluster 101 has lost one mobile device. In practical terms, this will be interpreted as one person that got off the vehicle (e.g. a bus).

[0089] The turnover can then be used to estimate the crowdedness of a cluster (e.g., corresponding to a vehicle) at specific tracts along an entire cluster path, e.g. vehicle path.

[0090] The data processing unit 1020 is configured to, based in part upon the turnover of the cluster, characterize a state of the cluster 101. The state of the cluster is determined based on the position, the velocity, and the number of mobile devices belonging to the cluster. This information is aggregated for all signals provided by the mobile devices that belong to the cluster in time.

[0091] More specifically, the state of the cluster may include, e.g.,: the life period of the cluster (i.e. the time interval in which the cluster is valid, defined for example in terms of a start time and an end time; the number of mobile devices belonging to the cluster (this parameter may vary over time, given to mobile devices that enter into I exit from the cluster after the latter has been recognized and defined, thus this parameter can be represented as a function with respect to time); the trajectory / path followed by the cluster during its life period.

[0092] The state of the cluster is then used to determine a set of time-based statistics for the cluster. The time periods of particular interest outside of the contemporary state include hours, days, weeks, months, and years. In the case of a vehicle, these statistics may preferably include a level of congestion aboard the same vehicle. The level of congestion aboard the vehicle may be an estimation of the total number of passengers based on the cluster knowledge of a subset of the total number of passengers.

[0093] Preferably, the time-based statistics may also breakdown the level of congestion along the specific tracts followed by the cluster or vehicle. The congestion along specific sub-tracts followed by the cluster 101 is particularly advantageous in its ability to define highly desired tracts.

[0094] The Applicant notes that, in the case of public transport, there are methods for determining where a user enters onboard the public transport network, but that many users are not adequately tracked using these conventional methods. On a bus, for example, many users with a monthly or yearly pass may avoid entirely tapping or checking onto the bus with their transit pass / pass. Additionally, when exiting the bus, there generally is no requirement to re-validate the transit pass or payment method to exit, thus lowering the quality of information obtained about the path taken by an individual.

[0095] In an embodiment, the method further comprises a step of identifying an additional cluster, generated and departing from the cluster 101 : this may happen, for example, if a relevant number of mobile devices (in particular a number equal or larger than the minimum cluster dimension), belonging to cluster 101, at a moment in time leave the latter and unitedly take a different direction, thereby creating a new cluster, distinct form cluster 101. The analysis on this additional cluster will then be continued in the same way disclosed herein for cluster 101.

[0096] The resulting data obtained from the clustering operation can be optionally presented for further analysis. In a particular embodiment, the data can be provided to end users via a dedicated portal. The dedicated portal may be provided by a server 131 and a second associated storage system 130 configured to store and serve the requested data. Section 1030 in figure 1 includes the server 131 and the second storage system 130.

[0097] Depending upon the end user's needs, this dedicated portal can be a website, a mobile application, or an API endpoint whereby users may request specific information associated with a specific tract or vehicle. The requested information may describe a corresponding state of the vehicle or a historical state of the vehicle according to the listed time periods of interest. An embodiment of the invention is shown in the method of figure 2. The process performed by the system of figure 1 comprises an initial step 2000 of indicating a geographic area of interest 1000 upon which to perform the analysis.

[0098] Step 2010 refers to an indication of the time period of interest. This time period may optionally contain only single traversals of the overall tract of interest or may extend over several years.

[0099] Step 2020 corresponds to the collection of report signals within the area of interest and time period defined in the previous steps.

[0100] Step 2030 corresponds to performing the clustering operation by the data processing unit 1020.

[0101] Optionally, Step 2040 corresponds to formatting and archiving the data from step 2030 for subsequent analysis. Preferably, step 2040 archives the data in either storage system 120 or 130, whereby diverse users may request the stored information. The stored information may then be provided through the dedicated portal 1030.

[0102] The invention achieves important advantages.

[0103] For example, the claimed method and system allow to generate interesting and useful outputs and statistics based on people transit, without violating any privacy rules or requiring any specific authorization from the users.

[0104] The characterization of transit can be useful for managing public transportation, in order to better meet the needs of customers (e.g. increasing vehicle availability at times and / or along paths for which usage is significantly higher than average).

[0105] A further advantage can be achieved through information services, which can make customers aware of crowdedness of certain vehicles and / or reliably inform regarding expected departure / arrival times.

[0106] Furthermore, statistics and trends concerning the usage of public transportation can be helpful in order to manage such services for environmental and sustainability purposes.

Claims

CLAIMS1. A method for the characterization of transit, comprising: determining an area of interest (1000); determining a time period of interest; obtaining, over time, report signals (106) provided by a plurality of mobile devices (102) within the area of interest (1000) in the time period of interest, each of said report signals (106) comprising a timestamp, a location parameter, and at least one of a temporary identifier and a session identifier; defining a cluster (101) of at least a subset of said plurality of mobile devices (102) based on the timestamp, the location parameter, and the at least one of a temporary identifier and a session identifier included in said report signals (106); determining, based on said report signals (106) obtained over time, a tract followed by the cluster (101); characterizing a state of said cluster (101) based on said report signals (106) obtained over time.

2. The method according to claim 1 wherein said report signals (106) further comprise signal characteristics; said signal characteristics comprising one or more of: a current service cell ID, a current service cell received power value, a current service cell received quality value, and a plurality of adjacent cell IDs, a plurality of adjacent cell received power values, and a plurality of adjacent cell received quality values.

3. Method according to claim 1 or 2 wherein said cluster (101) is further defined based on at least one of: a minimum persistence time of mobile devices in the cluster (101); a minimum cluster dimension; an entry / exit tolerance.

4. The method according to anyone of the preceding claims further comprising determining a turnover of said cluster (101), wherein said turnover is defined by a joining action and / or an exit action performed by one or more of said mobile devices (102).

5. The method according to claim 4 further comprising determining, based at least in part upon the turnover, a set of time-based statistics for said state of said cluster.

6. The method according to claim 5 wherein said set of time-based statistics comprises an average number of mobile devices in said cluster, said average number of mobile devices depicting said state of the cluster in a time period.

7. The method according to any one of the preceding claims wherein the report signals are sent from said mobile devices according to the Minimization of Drive Test, MDT, technology.

8. The method according to anyone of the preceding claims wherein said tract is at least a part of an overall path of a vehicle; the method further comprising characterizing a state of the vehicle based on the state of said cluster.

9. The method according to claim 8 wherein one of said plurality of mobile devices belongs to said vehicle defining a reference tract, said reference tract providing a reference point for defining said cluster.

10. The method according to claim 9 further comprising, based on said state of the vehicle, determining a set of time-based statistics describing a level of congestion aboard said vehicle.

11. The method according to claim 10 further comprising furnishing said set of time-based statistics to an end user to describe a corresponding current state and / or a historical state of said vehicle.

12. System for the characterization of transit, the system comprising a data processing unit configured to perform the method of any one of the preceding claims.

13. System according to claim 12 wherein said report signals are sent from said mobile devices according to the Minimization of Drive Test, MDT, technology.

14. System according to claim 13 wherein said tract is at least a part of an overall path of a vehicle; the data processing unit further configured to characterize a state of the vehicle based on the state of said cluster.

Images