WO2011158038A1

WO2011158038A1 - Tracking method

Info

Publication number: WO2011158038A1
Application number: PCT/GB2011/051134
Authority: WO
Inventors: Preet Khalsa; Valentin Mircea Burtea; Bernard Vincent Grush; Joseph Henry Leblanc; Kexing Rui
Original assignee: Skymeter Corporation; Skymeter Ltd
Priority date: 2010-06-17
Filing date: 2011-06-17
Publication date: 2011-12-22
Also published as: GB201010180D0

Abstract

A method of detecting whether a path has been followed by an object is described, comprising the steps of: receiving positioning data with respect to the position of the object at timed intervals; determining whether the object has followed the path by detecting a crossing, by the object, of at least one virtual gantry, the virtual gantry defined by a line segment in space crossing the path. Further optional validation checks for each gantry crossing are described, including the detection of the crossing of a number of virtual gantries, in an expected sequence, and over expected distances in an expected direction. A system for carrying out the method is also disclosed.

Description

TRACKING METHOD

The present invention relates to the field of global navigation system using satellites (GNSS). In particular, the invention relates to a method of more efficient and reliable tracking of the position of an object than is permitted by existing GNSS systems.

Satellite positioning systems and other types of wireless tracking technology can be used to determine the position of an object or a vehicle and to track progress of that vehicle, or other object, travelling along a path, generally along a road, or a public footpath, for example. The methods disclosed herein may be applied to any positioning system, although GNSS is used as a specific example to illustrate one particular application of the invention.

In some areas of commerce, it is desirable to be able to determine whether a vehicle has travelled along a particular road, path, or through a particular geographical area or zone of interest. This type of tracking is useful in the areas of road user charging, parking, insurance, car sharing schemes or any other use requiring metering based upon time, distance and geographical location.

It can be of interest to know whether a vehicle has travelled along a certain road segment, for example, when collecting tolls for that particular road, applying road pricing schemes, metering for the calculation of vehicle insurance risk. There exist a number of technologies available to determine the presence of a vehicle on a certain road segment.

Toll booths with human operators charging a fee for road use can be used. This approach can impair traffic flow, which may result in expensive and inconvenient traffic delays.

Automatic toll gantries may be used. In these systems, the vehicle is identified using a combination of the following: inductive loops or radar technology used to detect the presence of a vehicle; video capture for accurate vehicle classification (i.e. size, usage type); digital image capture and character recognition software for license plate identification; radio-frequency identification (RFID) or dedicated short-range communications (DSRC), which may be used to carry out an electronic transaction. In this electronic transaction, an RFID tag may be associated with a particular user or vehicle, which is in turn associated with particular account details held by the organisation responsible for collecting payment on the particular road segment. Recognition of the RFID tag in a particular zone will then result in the account of the vehicle user being charged for use of that road segment.

The above-described technologies can allow physical gantry systems to be automated. These automated gantries can alleviate the draw-back of human operated toll booths, which can result in traffic jams. However, they come with a significant installation and maintenance cost. Significant amounts of infrastructure and construction works must be carried out to put these systems in place and these can result in a high cost of deployment. A further drawback can be that physical gantries, be they automatic or manually operated, require a significant amount of space for installation and often cause temporary traffic interruption.

The use of GNSS technology to monitor whether a road segment has been used by a particular vehicle can benefit from an existing GNSS infrastructure, which has already been deployed. This can reduce the system deployment cost as compared to the cost of the above described physical installations.

However, a GNSS receiver is only able to determine the location of the person, object, or vehicle to which it is mounted with a finite degree of accuracy. Further, the performance of a GNSS receiver depends upon atmospheric conditions and the environment in which it is located. In dense urban areas the accuracy may degrade significantly due to signal obstruction by surrounding buildings, or due to multi-path reflections from surrounding buildings and objects. Existing GNSS systems use map-matching to try to compensate for inherent GNSS errors. This is where the calculated GNSS location is compared with a map stored in the GNSS receiver. This map-matching constrains the reported trajectory of the receiver to the existing roads or pathways registered on the map. Therefore, identifying whether the vehicle is present on a certain road segment would appear to become trivial. However, map-matching algorithms are not fool-proof. If the positional error induced in a receiver by environmental conditions or multi-path reflections exceeds the distance between two adjacent roads on a map, then resulting map registration errors may amplify the GNSS positioning errors rather than compensating for them. The receiver may therefore be reported as being located on the road adjacent to the road upon which it is actually located. This is of particular risk in the case of neighbouring or parallel roads. Furthermore, map-matching requires a significant degree of processing within the receiver and can increase operation and logistics costs. Methods of determining toll charges for vehicles using a traffic route are discussed in US 5717389. This document describes a road toll solution which does not rely on specific details of navigation maps. The concept behind the disclosure in US 5717389 is to validate a certain predefined trajectory by placing validation points along a route. The validation points have a circle of detection of a predefined radius. Although specified as a circle, these validation points can be any polygon. If the GNSS receiver resolves the position of the vehicle to be within one of these validation circles and if the vehicle passes through all of the validation circles, then it is concluded that the vehicle has travelled along the given segment of road. All of the validation areas in US 5717389 are generally overlapping circles and the processing required in verifying whether the GNSS receiver is within any of these circles can be burdensome on the device carrying out this processing.

The present invention attempts to overcome the draw-backs of all of the above technologies, by providing a GNSS-based road use monitoring method which requires a minimum of data storage and requires a minimum of data processing in order to determine whether a section of road or path or some other geographical zone has been travelled through by a user.

According to the present invention there is provided a method of detecting whether a path has been followed by an object,

comprising the steps of:

receiving positioning data with respect to the position of the object in timed intervals;

determining whether the object has followed the path by detecting a crossing, by the object, of at least one virtual gantry, the virtual gantry defined by a line segment in space crossing the path. The method according to the present invention is based upon the detection of the object having crossed a single line segment in space. The detection of this line crossing requires minimal processing power and the storage of the series of line segments relating to virtual gantry positions makes efficient use of memory storage to achieve the object of detecting whether the object has followed the path. The method of the present invention removes the need for any processing to calculate whether the tracked object is on any part of a predefined map of roads and removes the need for a processor determine whether an object is within a particular polygonal area.

Determining whether the object has followed the predetermined path may further comprise the step of detecting a crossing, by the object, of a plurality of virtual gantries located at intervals along the path. Detecting a plurality of virtual gantries located at intervals along the path can further improve the reliability and accuracy of the method of the present invention. It can help to compensate for 'phantom' gantry crossings, where inaccuracy in a GNSS calculated position may give a false indication of a virtual gantry located near to a neighbouring path having been crossed, when the object was in fact moving along the neighbouring path and not along the path to which the virtual gantry relates.

Detecting a crossing of a virtual gantry may further comprise the step of detecting a segment between two consecutive object positions intersecting a virtual gantry. The detecting of a line segment drawn between two consecutive calculated positions of the object intersecting the line segment representing a virtual gantry is a simple and efficient processing step and provides a simpler zone detection method when compared to other known methods, including calculation of whether an object is within any particular circular, polygonal or other 2D area.

Determining whether the object has followed the path may further comprise the step of detecting the direction in which a virtual gantry has been crossed by the object. Detection of the direction in which the virtual gantry has been crossed can allow the verification of whether the calculated direction of travel of the object correctly corresponds to the expected direction of travel of the object along the path and across the virtual gantry. This further provides a validation check for a calculated use of a path. Determining whether the object has followed the path may further comprise the step of comparing whether the detected virtual gantry crossing relates to a virtual gantry different from the virtual gantry detected in the previously detected gantry crossing. This step can provide a further validation check to verify that a detected gantry crossing is not merely due to jitter and/or inaccuracy in the calculated position of the object.

The distance actually travelled by the object between virtual gantry crossings may be measured and compared with a distance along the predetermined path. The measurement of the distance may be calculated from the GNSS signal or from another input means, such as an odometer in a vehicle or any other distance measuring device. A comparison of this measured distance travelled by the tracked object with an expected distance that an object should travel between two virtual gantries, if travelling along the predetermined path, can provide a further validation check for the detection of an object having travelled along the predetermined path.

The path followed by the object may only be defined as the predetermined path when the distance actually travelled by the object is within a predefined range. The predefined range may be a range above and/or below the actual distance along the predefined path between the virtual gantries. The use of a range can allow for an inaccuracy in the measurement of the distance that the object has travelled in between gantry crossings and the range may be chosen to be large enough to allow for inaccuracies in measurement of the distance, but small enough that an object taking a second, different path between virtual gantry crossings is not recorded. The predefined range may be defined relative to a distance along the predetermined path.

Detected gantry crossings may be compared with any one of a NOT, AND or OR relationship between virtual gantries. Applying this further logical comparison to detected gantry crossings can improve the reliability of the detection of genuine path usage.

The method may further comprise the step of recording on a database a value indicating that the object has travelled along the predefined path. This step allows the user to access the memory to extract a list of predefined paths along which the object has travelled. The recording step may only be carried out if the object has crossed the gantry in an expected direction. The recording step may only be carried out if the object has crossed a plurality of gantries in an expected sequence. The recording step may only be carried out if the most recently detected virtual gantry crossing relates to a virtual gantry different from the virtual gantry detected in the previously detected gantry crossing. The recording step may only be carried out if a detected crossing complies with any one, or a combination of, a NOT, AND or OR relationship between virtual gantries. Applying these conditions can result in a more reliable recording of the objects having travelled along the predetermined path, since a greater number of validation checks must be successfully passed before the recording step is carried out.

The invention further provides a system for detecting whether a path has been followed by an object, comprising:

a receiver for receiving positioning signals from positioning satellites; a processing unit for processing position data generated by the receiver; a first memory for storing data representing the positions of virtual gantries; and

an output memory for recording paths which an object is determined to have followed;

wherein the system is configured to carry out any one of the methods described herein.

A particular embodiment of the invention will now be described in further detail with reference to the following figures, in which:

Figure 1 shows an example of a road system and corresponding virtual gantries; and

Figure 2 shows a schematic drawing of a system which may be used to implement the method of the present invention.

In the Figures, numerals beginning with the number 1 indicate real-world objects, while numerals beginning with 2 indicate virtual lines or areas in space. Numerals beginning with the number 3 relate to system components of the system carrying out the methods of the present invention.

Figure 1 shows a road system 10 comprising a toll road 100 and a free road 101. Superimposed on the road system is plurality of line segments, representing virtual gantries 201 , 202 and 203. An object 102 is travelling along toll road 100. A second object 103 is travelling along free road 101 in a first direction. A third object 104 is travelling along the same free road 101 in a second direction. Each of objects 102, 103 and 104 is equipped with a GNSS receiver or similar tracking device. Each of the tracking devices will receive signals from a global positioning system 105, to determine their respective positions.

The application of the method of the present invention will now be described in relation to the particular cases of moving objects 102, 103 and 104 as they follow the trajectories depicted in Figure 1.

Object 102, which may be any object or vehicle carrying a GNSS receiver, is travelling along toll road 100 and, as can be seen from the figure, its trajectory will pass across virtual gantries 201 , 202 and 203. As the object passes along the road, it will be expected that the object passes the virtual gantries in a particular sequence, i.e. in this case, virtual gantry 201 followed by virtual gantry 202, followed by virtual gantry 203. A sequence in which gantry crossings are detected can be used to further verify whether a detected gantry crossing is a genuine virtual gantry crossing or a "phantom" gantry crossing caused by jitter or inaccuracy in the GNSS signal. There may be sufficient jitter in the GNSS signal and resultant calculated position to bring about multiple calculated crossings of one of virtual gantries 201 , 202 or 203. As object 102 crosses the position of virtual gantry 201 , the GNSS calculated position may have sufficient jitter that the object appears to cross backwards behind the virtual gantry and then crosses again forwards. Depending upon the amount of jitter and speed of the object, this may happen a number of times. The method of the present invention may optionally include comparing the position of the most recently detected gantry crossing with the position of the previously detected gantry crossing. If these relate to the same virtual gantry, then the most recent gantry crossing may be disregarded. If, however, the two last detected gantry crossings relate to different virtual gantries, then they are more likely to be genuine gantry crossings. Further, the sequence in which the virtual gantries are apparently crossed by the object can be compared with an expected sequence in which the gantries would normally be crossed as object 102 travels along toll road 100. Further, a distance along toll road 100 will necessarily be travelled by an object if it is actually following the troll road in between virtual gantry crossings. For example, if the distance along the toll road between virtual gantry 201 and virtual gantry 202 is known, then this distance may be compared with the actual distance travelled by the vehicle during the time that has elapsed between the last detected virtual gantry crossing of virtual gantry 201 and the detected gantry crossing of gantry 202. This known expected distance may be compared with an actual distance travelled. The actual distance travelled may be taken from another system in the object, for example, if the object is a car, the actual distance travelled by the car may be taken from its odometer or other on board system, and compared with the expected distance between gantry crossings.

Due to jitter and inaccuracy in the GNSS signal, along with the range of actual trajectories that an object may follow along toll road 100, a range of distances may be covered by the object as it passes between the position of virtual gantry 201 and virtual gantry 202 along the toll road 100. Therefore, it is advantageous to make the comparison between the actual distance travelled by the car and a potential range of distances that the car may have travelled in between virtual gantry crossings.

Object 103, which may be any object or vehicle carrying a GNSS receiver, is travelling along free road 101. The vehicle is travelling generally left to right along the free road 101 as shown by arrow 105 in the figure. The GNSS signal associated with object 103 is subject to a certain degree of jitter and inaccuracy, which results in a calculated position for the object which may be anywhere approximately within circle 204. As the object travels along the free road, the object will pass adjacent to virtual gantry 202. If there is sufficient jitter in the signal and the object 103 passes sufficiently close to the virtual gantry 202, then a crossing of that virtual gantry by object 103 may be detected. However, the object not having crossed virtual gantries 201 or 203, it is possible to conclude that the object was not passing along the toll road 100 and thus it will not be recorded that the object has passed along the toll road in this particular case.

Object 104 is travelling along the free road in the opposite direction to object 103, generally right to left. An equivalent amount of jitter and potential inaccuracy in the calculated position of object 104 from the GNSS signals will apply, as depicted in circle 204 for object 103. As object 104 passes virtual gantry 202, again, jitter or inaccuracy in its calculated position may cause its calculated trajectory to appear to cross virtual gantry 202, However, this is likely to be in a direction opposite to the expected direction of travel for toll road 101 at the position of gantry 202. This expected direction is indicated by arrow 205. It is still possible that sufficient jitter in the GNSS signal, or inaccuracy caused by signal reflections, could indicate that object 104 has crossed gantry 202 in the direction of arrow 205. Again, in this case, the object not having passed through virtual gantries 201 and 203, it can be concluded that object 104 has also not travelled along the toll road 100. In some cases, it may already be known that the distance between toll road 100 and free road 101 is sufficient that jitter or inaccuracy in a calculated GNSS position will not be sufficient to give a false indication that either of objects 103 or 104 has crossed virtual gantry 202. In this case, it will not be necessary to detect whether or not either of gantries 201 or 203 has been crossed to determine whether the objects have passed along the toll road or not. Where a virtual gantry is sufficiently distant from any neighbouring roads, a simple detection of a crossing of a single gantry may be sufficient to give a reliable indication of whether a section of toll road has been used or not. It will therefore be apparent that, for example, the use of a single gantry 201 may be sufficient to reliably detect whether an object has passed along toll road 100. Similarly, detection of a crossing of gantry 203 may also be sufficient. Where this certainty is not possible due, for example, to multiple adjacent paths, then further validation checks may be necessary to validate whether a detected gantry crossing is genuine or not.

In the event of a temporary loss of GNSS signal, or if a section of road is particularly susceptible to inaccuracies in GNSS positioning calculations, perhaps due to surrounding buildings, plant life or geographical variations such as hills or mountains, it may be preferable to use a number of virtual gantry points along the toll road. Using a greater number of virtual gantries may improve the reliability of determining whether the toll road has been used or not. Furthermore, requiring a minimum number of virtual gantries on the toll road to be crossed before deciding that the toll road has indeed been used can render the method even more reliable. Further relationships may be defined between gantries, in the form or logical OR, AND and NOT relationships. In the example illustrated in Figure 1 , some example relationships may be as follows. An AND relationship may be recorded between, for example, virtual gantry 201 and 202. In this case, it will only be determined that the object 102 has travelled along path 100 if crossings of both of the virtual gantries 201 AND 202 have been crossed. This AND relationship may be applied to any pairing, or all three, of virtual gantries 201 , 202 and 203. An OR relationship may further be applied to a pair or plurality of virtual gantries. For example, it may be determined that the object 102 has travelled along path 100 if crossings of either of virtual gantries 201 OR 203 has been crossed. A NOT relationship may also be applied to certain virtual gantries. This may be applied where an object is only deemed to have travelled along a path if a detection of a crossing of a particular virtual gantry has not been detected. This could apply to a case in Figure 1 , where a virtual gantry 206 is optionally located on path 101. The virtual gantry 206 may be located away from path 100, for example, approximately at the location of either of objects 103 or 104 in the figure, so as to make phantom virtual gantry crossings due to signal inaccuracies or signal jitter less likely. Path 100 may only be determined to have been followed by object 102, if a crossing of the virtual gantry on path 101 has NOT been detected. This NOT relationship may be applied as an alternative to, or in combination with, either of the above described AND and OR relationships. For example, path 101 may only be deemed to have been travelled along by object 102 if crossings of virtual gantries 201 AND (202 OR 203) NOT 206, have been detected.

From the above description, it will be apparent that, by using the method of the present invention, the passage of a vehicle through a set of virtual gantries in a predetermined sequence may be considered as sufficient evidence that the vehicle has travelled the entire length of a given road section. The virtual gantries are segments of a certain width intersecting the trajectory followed by the road. The width of the virtual gantry may be slightly wider than the physical road which it is placed to intersect. This ensures that the risk of jitter or inaccuracy in GNSS position calculations does not cause the calculated trajectory of the car to pass around the edge of the gantry. This gives greater certainty of an object crossing the virtual gantries positioned on the path being detected. The gantry crossing is recorded if the signal between two consecutive vehicle positions intersects the signal representing the virtual gantry. The gantry may be sensitive to one direction only, in that if the detected gantry crossing is not in an expected direction for the virtual gantry, then the gantry crossing may be ignored. Multiple subsequent crossings of the same virtual gantry may be ignored, since these may be a consequence of jitter due to GNSS errors or inaccuracies. It may be required that all gantries are triggered in a predetermined sequence to indicate that the vehicle has travelled along the particular road segment corresponding to that set of virtual gantries. A distance travelled between detected gantry crossings may further be required to be within a predefined range of expectation for a gantry crossing to be recorded and use of a toll road to be consequently recorded.

Figure 2 is a schematic representation of a system which may be used to implement the method of the present invention. A GNSS receiver 300 is configured to receive GNSS positioning signals from GNSS satellites 105 of Figure 1. The position data generated by the receiver is passed via a communication line 301 to a processing unit. The passing of the information between the GNSS receiver and the processor may be done in real time as the position is calculated. Alternatively, the data may be saved into an intermediate memory and accessed by a processor carrying out the method of the present invention at a later point in time. Processor 303 is connected to an input memory containing data representing the positions of the virtual gantries and to an output memory for recording at least which toll roads have been used, along with the output of any of the comparison steps, for example any gantry systems that have been detected. A first memory 304 may contain the positions of the virtual gantries and the second memory 305 may contain the output data in a suitable format. Processor 303 monitors the incoming GNSS position data and checks for a line segment in-between two sequential object positions crossing a virtual gantry segment held in first memory 304. This check can be carried out using any known algorithm for the detection of line intersections. Processor 303 may be programmed to subsequently carry out any of the validation checks on that gantry crossing, for example comparing a detected sequence in which gantries have been crossed with an expected sequence of gantry crossings, comparing the direction in which a gantry has been crossed with an expected direction. The expected results of these validation checks may be stored in memory 304 along with the position data for each gantry. The processor may further compare a recorded distance since the previous gantry crossing with an expected distance between the two relevant virtual gantries.

The memory 304 may contain at least the end point coordinates of each gantry and for each virtual gantry. Moreover, there may also be stored any or all of: an expected crossing direction; an expected previous gantry crossing; an expected subsequent gantry crossing; an expected distance since a previous gantry crossing; an identifier corresponding to the toll road to which that virtual gantry corresponds; and information concerning any NOT, AND or OR relationships for each gantry.

The output into memory 305 may contain at least a record of the toll road that the object has deemed to have travelled along, together with, optionally, each detected virtual gantry crossing, each validated gantry crossing and the times and distances corresponding to those gantry crossings.

It will be apparent to the skilled reader that the advantages of the described system are a simple detection and validation method, requiring low memory storage and low processor loading. The use of unidirectional gates may deal with the problem of GNSS system jitter, as may ignoring an indication that the object has passed through the same gantry several times within a set time period. Repeated crossings may be recorded if they are sufficiently far apart in time. Less specific tuning of each virtual gantry size or position is required as compared to prior art systems. It is not necessary to select the radius of every detection point to find a balance between the distances between the two parallel running roads, balanced with the magnitude of potential GNSS errors. The selection of virtual gantry positions away from the areas of parallel road, in combination with the described validation checks on each gantry crossing can allow the virtual gantry dimensions for each virtual gantry recorded in the system to be standardised to a greater degree.

A benefit of the described method is that the trajectory of the vehicle is indirectly constrained by the road network to which the virtual gantries correspond, even though road mapping information is not a direct part of the detection process. By the use of virtual gantries in areas where likelihood of road confusion is reduced, sufficient validation of the entire trajectory of the object may be gained without the need for constant comparison with actual road maps and without the need for high numbers of validation points corresponding to a road section. The described system does not require the tracking of every point of the object along an entire road section, but the analysis is focussed on a relatively simple detection method implemented at certain key locations on the road network.

Claims

1. A method of detecting whether a path has been followed by an object, comprising the steps of:

determining whether the object has followed the path by detecting a crossing, by the object, of at least one virtual gantry, the virtual gantry defined by a line segment in space crossing the path.

2. A method according to claim 1 , wherein determining whether the object has followed the path further comprises the step of:

detecting a crossing, by the object, of a plurality of virtual gantries located at intervals along the path.

3. A method according to claim 1 or claim 2, wherein detecting a crossing of a virtual gantry further comprises the step of:

detecting a segment between two consecutive object positions intersecting a virtual gantry.

4. A method according to any one of the preceding claims, wherein determining whether the object has followed the path further comprises the step of:

detecting the direction in which a virtual gantry has been crossed by the object.

5. A method according to any one of the preceding claims, wherein determining whether the object has followed the path further comprises the step of:

comparing whether the detected virtual gantry crossing relates to a virtual gantry different from the virtual gantry detected in the previously detected gantry crossing.

6. A method according to any one of the preceding claims, wherein the distance actually travelled by the object between virtual gantry crossings is measured and compared with a distance along the predetermined path.

7. A method according to claim 6, wherein the path followed by the object is only defined as the predetermined path when the distance actually travelled by the object is within a predefined range.

8. A method according to claim 7, wherein the predefined range is defined relative to a distance along the predetermined path.

9. A method according to any one of the preceding claims, wherein detected gantry crossings are compared with any one of a NOT, AND or OR relationship between virtual gantries.

10. A method according to any one of the previous claims, further comprising the step of recording in a memory a value indicating that the object has travelled along the predetermined path.

1 1. A method according to claim 10, wherein the recording step is only carried out if the object has crossed the virtual gantry in an expected direction.

12. A method according to claim 10, wherein the recording step is only carried out if the object has crossed a plurality of gantries in an expected sequence.

13. A method according to claim 10, wherein the recording step is only carried out if the most recently detected virtual gantry crossing relates to a virtual gantry different from the virtual gantry detected in the previously detected gantry crossing.

14. A method according to claim 10, wherein the recording step is only carried out if a detected series of virtual gantry crossings complies with any one, or a combination of a NOT, AND or OR relationship between virtual gantries.

15. A system for detecting whether a path has been followed by an object comprising:

wherein the system is configured to carry out the method of any one of claims 1 to 14.