Foreign object detection for railway

Abstract

According to a first aspect of the disclosure, there is provided a collision avoidance system for an on-track plant vehicle, comprising: an object detecting system configured to detect an object within an area of interest surrounding the vehicle; a trajectory determining system configured to determine a trajectory of the vehicle; a collision determining system configured to determine a probability of collision occurrence based on an output of the object detection system and the trajectory determining system.

Description

[0001] FOREIGN OBJECT DETECTION FOR RAILWAY

[0002] TECHNICAL FIELD

[0003] The present disclosure relates to a collision avoidance system and method of collision avoidance.

[0004] BACKGROUND ART

[0005] Maintenance of railway infrastructure is often carried out using on-track plant (OTP). OTP is machinery designed to run on the railway. OTP may include rail mounted maintenance machines (RMMMs) and road rail vehicles (RRVs) for example, and any trailers and attachments. RRVs may be driven on roads as well as the railway. RRVs may, for example, comprise flanged wheels suitable for travelling on rail tracks that may be deployed for rail travel and retracted for road travel. Fig. 1. shows an example RRV.

[0006] OTP has limited manoeuvrability as it is restricted by the tracks on which it travels. Additionally, the visibility of a driver of OTP may be limited, for example by parts of the OTP itself or its cargo. This increases the likelihood of collisions with the OTP, which may cause damage to equipment, or injury to people.

[0007] Patent publication GB2595747 A discloses an anti-collision apparatus for RRVs that detects objects that might present a collision risk. The apparatus disclosed will be limited in its effectiveness at speeds higher than around 5 mph during transit. Limiting factors include the radar sensor range of 10 m, which is not sufficient to meet stopping distances set out in BS EN 15746-2:2020. Trailers and attachments ahead of sensor mounting also extend the range required, as will any response time allowed for the operator to react to a warning prior to any automated braking.

[0008] The present disclosure aims to at least partially solve some of the above problems. It should be noted that, typically, such an apparatus is primarily (but not solely) used when the vehicle is in transit to or from a work site, rather than while working at the work site itself. SUMMARY OF THE INVENTION

[0009] According to a first aspect of the disclosure, there is provided a collision avoidance system for an on-track plant vehicle, comprising: an object detecting system configured to detect an object within an area of interest surrounding the vehicle; a trajectory determining system configured to determine a trajectory of the vehicle; a collision determining system configured to determine a probability of collision occurrence based on an output of the object detection system and the trajectory determining system.

[0010] Optionally, the system further comprises a collision alert system configured to generate an alert based on the determination by the collision determining system.

[0011] Optionally, the system further comprises an automatic braking system configured to automatically slow the vehicle based on the determination by the collision determining system.

[0012] Optionally, the collision determining system is configured to assign an object detected by the object detecting system a probability of collision occurrence based on an intersection between the location of the detected object and the determined trajectory of the vehicle, a higher probability being assigned when there is intersection between the location of the detected object and the determined trajectory, and a lower probability being assigned when there is no intersection between the location of the detected object and the determined trajectory of the vehicle.

[0013] Optionally, the object detection system is configured to determine a trajectory of a detected object that is moving; and the collision determining system is configured to assign an object detected by the object detecting system a probability of collision occurrence based on an intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle, a higher probability being assigned when there is intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle, and a lower probability being assigned when there is no intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle. Optionally, a higher probability is assigned when a position of the object on its determined trajectory intersects a position of the vehicle on its determined trajectory at the same point in time, and a lower probability is assigned when a position of the object on its determined trajectory does not intersect a position of the vehicle on its determined trajectory at any point in time.

[0014] Optionally, the object detecting system is configured to determine whether a detected object is a known type of object; and the collision determining system is configured to assign a probability of collision occurrence based on whether a detected object is determined to be a known type of object.

[0015] Optionally, the object detecting system comprises at least one transmitter configured to transmit a signal and at least one corresponding receiver configured to receive a signal reflected from an object. Optionally, the at least one transmitter and at least one receiver comprises a Lidar transmitter and receiver. . Alternatively, or additionally, the object detecting system comprises at least one video camera.

[0016] Optionally, the trajectory determining system is configured to combine information on vehicle motion and position with detected potential routes along the railway in the direction of motion, to provide forward predictions of the vehicle trajectory.

[0017] Optionally, the trajectory determining system comprises a rail detecting system, a motion detecting system and / or a location detecting system. Optionally, the trajectory determining system further comprises a route determining system configured to determine a route followed by the vehicle based on map information.

[0018] The rail detecting system is configured to detect rails on which the vehicle travels and determine the path formed by the rails. Optionally, the rail detecting system comprises at least one camera configured to image the rails.

[0019] The motion detecting system is configured to determine parameters relating to the motion of the vehicle. Optionally, the parameters include one or more of: direction of travel, speed, acceleration. Optionally, the motion detecting system comprises an accelerometer. Alternatively, or additionally, the motion detecting system may be configured to be derived from tracking of object detections over time. Optionally, the motion detecting system may be configured to process external inputs from the vehicle itself, such as vehicle speed.

[0020] The location detecting system is configured to determine a location of the vehicle.

[0021] Optionally, the location detecting system is a satellite positioning system, such as the Global Positioning System (GPS).

[0022] Optionally, the system comprises one or more processors configured to perform data processing functions of each of the object detecting system, the trajectory determining system and the collision determining system.

[0023] In a specific example, the system further comprises a collision alert system configured to generate an alert based on the determination by the collision determining system In the specific example, the collision determining system is configured to assign an object detected by the object detecting system a probability of collision occurrence based on an intersection between the location of the detected object and the determined trajectory of the vehicle, a higher probability being assigned when there is intersection between the location of the detected object and the determined trajectory, and a lower probability being assigned when there is no intersection between the location of the detected object and the determined trajectory of the vehicle. In the specific example, the object detection system is configured to determine a trajectory of a detected object that is moving. In the specific example, the collision determining system is configured to assign an object detected by the object detecting system a probability of collision occurrence based on an intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle, a higher probability being assigned when there is intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle, and a lower probability being assigned when there is no intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle. In the specific example, the object detecting system comprises at least one Lidar transmitter configured to transmit a signal and at least one corresponding Lidar receiver configured to receive a signal reflected from an object. I the specific example, the trajectory determining system comprises a rail detecting system, a motion detecting system and a location detecting system. The rail detecting system is configured to detect rails on which the vehicle travels and determine the path formed by the rails. The rail detecting system comprises at least one camera configured to image the rails. The motion detecting system is configured to determine parameters relating to the motion of the vehicle. The parameters include one or more of: direction of travel, speed, acceleration. The motion detecting system comprises an accelerometer and is configured to determine motion additionally based on tracking of object detections over time.

[0024] According to a second aspect of the disclosure there is provided a method of collision avoidance for an on-track plant vehicle, comprising: detecting an object within an area of interest surrounding the vehicle; determining a trajectory of the vehicle; determining a probability of collision occurrence based on an output of the object detection and the trajectory determination.

[0025] BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Further features of the disclosure will be described below, by way of non-limiting examples and with reference to the accompanying drawings, in which:

[0027] Fig. 1 shows an example road rail vehicle;

[0028] Fig. 2 shows a schematic of an example functional configuration of a collision avoidance system;

[0029] Fig. 3 shows a first example of the collision avoidance system in operation;

[0030] Fig. 4 shows a second example of the collision avoidance system in operation; and Fig. 5 schematically shows an example hardware configuration of the collision avoidance system.

[0031] DETAILED DESCRIPTION

[0032] Fig. 2 schematically shows an example functional configuration of a collision avoidance system 1 for an on- track plant (OTP) vehicle, such as an RRV as show in Fig. 1. As shown, the collision avoidance system 1 comprises an object detecting system 2, a trajectory determining system 3, and a collision determining system 4. As shown, the collision avoidance system 1 may further comprise a collision alert system 5, and / or an automatic braking system 6. The purpose of the collision avoidance system is to avoid collisions between the vehicle and an object. The collision avoidance system 1 may therefore detect potential collisions and may then take mitigating action, such as raising an alarm or slowing or stopping the vehicle.

[0033] The collision determining system 4 is configured to determine a probability of collision occurrence based on an output of the object detection system 2 and the trajectory determining system 3. Combining the outputs of the object detection system 2 and the trajectory determining system 3 in this way will allow more accurate prediction of collisions and reduce false positives, e.g. as the requirements for detection distances increase with vehicle speed. The collision determining system 4 may be implemented by one or more processors (e.g. executing software code), configured to process information from the object detection system 2 and the trajectory determining system 3.

[0034] The collision alert system 5 is configured to generate an alert based on the determination by the collision determining system. For example, when the probability of a collision is over a predetermined threshold and / or within a predetermined time to collision, the collision alert system 5 may be configured to provide a visual and / or audible alert. A visual alert may be a flashing light, for example. An audible alert may be a siren, for example.

[0035] The automatic braking system 6 may be configured to automatically slow the vehicle based on the determination by the collision determining system. For example, when the probability of a collision is over a predetermined threshold, the automatic braking system 6 may be configured to engage a braking system of the vehicle to slow the vehicle (e.g. to a stop) to attempt to avoid a collision or mitigate damage. The automatic braking system 6 may be operatively connected to the vehicle braking system, for example, and may provide control signals to the vehicle braking system.

[0036] The object detecting system 2 is configured to detect an object within an area of interest surrounding the vehicle. The area of interest may be a predefined area that is considered to be at risk of collision. This area may preferably include the three-dimensional footprint of the vehicle extended to cover potential trajectories over the stopping time required for the current vehicle speed, and / or may include a further region that is a predetermined distance from the vehicle to enable detection and tracking of moving obstacles over a wider range. The area of interest may depend on the specific type of vehicle. The area of interest may be monitored by one or more sensing systems that form part of the object detecting system 2. The sensing system of the object detecting system 2 preferably comprises at least one Lidar transmitter configured to transmit a signal and at least one corresponding Lidar receiver configured to receive a signal reflected from an object. The presence, and preferably three-dimensional location, of an object within the area of interest may be determined based on the signal received by the receiver. The signal may be analysed by one or more processors forming part of the object detecting system 2. The Lidar is preferably capable of object detection at ranges exceeding that required for vehicle stopping distances (e.g. as defined in BS EN 15746) whilst also allowing time for the system to respond and the operator to react to a warning. Additionally, the resolution, beam divergence and coverage of the Lidar is preferably sufficient to ensure objects can be reliably and accurately located at these ranges, within a cluttered railway environment - a challenge that becomes increasing difficult as range increases.

[0037] The sensing system of the object detecting system 2 may comprise at least one camera, e.g. video camera (i.e. may be a camera sensing system). The presence of an object within the area of interest may be determined based on the images captured by the camera. The images may be analysed by one or more processors forming part of the object detecting system 2.

[0038] The trajectory determining system 3 is configured to determine a trajectory of the vehicle. Information about the trajectory of the vehicle may be used to improve the accuracy of the collision avoidance system, for example by assigning a higher probability of collision to objects that the vehicle is moving towards than objects that the vehicle is moving away from.

[0039] In one specific example system, the trajectory determining system 3 may comprise a rail detecting system configured to detect rails on which the vehicle travels and determine a path formed by the rails. This path may be used to determine the trajectory of the vehicle. The rail detecting system may comprise at least one camera configured to image the rails. Based on the image data the trajectory determining system may perform image processing to detect the rails and determine the trajectory. In another specific example, the trajectory determining system may comprise a route determining system configured to determine a route followed by the vehicle based on map information. This route may be used to determine the trajectory of the vehicle.

[0040] In another specific example the trajectory determining system 3 may comprise both a rail detecting system and a route determining system.

[0041] In another specific example system, the trajectory determining system may comprise a motion detecting system configured to determine parameters relating to the motion of the vehicle. The motion detecting system may be used in combination with a rail detecting system and / or a route determining system. The parameters may include one or more of: direction of travel, speed, and acceleration. This may include linear and / or angular speed and / or acceleration. For example, the motion detecting system may comprise an accelerometer. The motion detecting system may, alternatively or additionally, be configured to process inputs external to the system, e.g. from the vehicle itself, such as speed. Such inputs may be provided through an interface with the vehicle CAN bus, for example.

[0042] In another specific example system, the trajectory determining system may comprise a location detecting system configured to determine a location of the vehicle. The location detecting system is a satellite positioning system, for example, such as the Global Positioning System (GPS). The location detecting system may be used in combination with a rail detecting system, a route determining system, and / or motion detecting system.

[0043] Fig. 3 shows a first illustrative example of the collision avoidance system in operation. Fig. 3 shows an environment in which the vehicle is operating, including the track and two detected objects (boxed by dashed lines). The dot-dashed area denotes a trajectory of the vehicle, with the arrow in the Figure indicating the direction of travel on the trajectory. Each detected object has a probability of collision assigned (P=X and P=Y). In this case, P=X indicates a high probability of collision, denoted by the hash within the dashed lines.

[0044] As illustrated in Fig. 3 the collision determining system may be configured to assign an object detected by the object detecting system a probability of collision occurrence based on an intersection between the location of the detected object and the determined trajectory of the vehicle. A higher probability may be assigned when there is intersection between the location of the detected object and the determined trajectory, and a lower probability may be assigned when there is no intersection between the location of the detected object and the determined trajectory of the vehicle. P=X is a higher probability of collision than P=Y in Fig. 3. Assigning a probability in this way may help to improve the accuracy of the system and reduce false positives.

[0045] Fig. 4 shows a second illustrative example of the collision avoidance system in operation. Fig. 4 also shows an environment in which the vehicle is operating, including the track and two detected objects (boxed by dashed lines). Again, the dot-dashed area denotes a trajectory of the vehicle, with the arrow in the Figure indicating the direction of travel on the trajectory. Additionally, in Fig. 2 the objects are moving and the trajectory of the objects is determined by the object detecting system 2. The trajectory is illustrated by the further dashed lines and additional arrows showing the direction of travel of the objects. Again, each detected object has a probability of collision assigned (P=X and P=Y). In this case, P=Y indicates a high probability of collision, denoted by the hash within the dashed lines.

[0046] As illustrated in Fig. 4 the object detection system may be configured to determine a trajectory of a detected object that is moving. The collision determining system may be configured to assign an object detected by the object detecting system a probability of collision occurrence based on an intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle. A higher probability may be assigned when there is intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle, and a lower probability being assigned when there is no intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle. P=Y is a higher probability of collision than P=Y in Fig. 3. Assigning a probability in this way may help to further improve the accuracy of the system and further reduce false positives. This also allows the system to detect where vehicles are in convoy during transit, and alarm only when distance limits for this mode of transit are broken.

[0047] In some example systems, a higher probability may be assigned when a position of the object on its determined trajectory intersects a position of the vehicle on its determined trajectory at the same point in time, and lower probability may be assigned when a position of the object on its determined trajectory does not intersect a position of the vehicle on its determined trajectory at any point in time. Assigning a probability in this way may help to still further improve the accuracy of the system and still further reduce false positives.

[0048] In some example systems, the object detecting system 2 may be configured to determine whether a detected object is a known type of object. The collision determining system 4 may be configured to assign a probability of collision occurrence based on the whether a detected object is determined to be a known type of object. For example, the collision determining system 4 may reduce the probability assigned to objects such as tunnels, or rail infrastructure that may be inherently low risk. On the other hand, the collision determining system 4 may increase the probability assigned to objects such as people that may be inherently high risk.

[0049] In any of the above examples, uncertainty in the determined location or trajectory of the object or vehicle may be factored into the assigned probability. Further, probabilities may be divided into a plurality of levels, such as high, medium and low. Each of these may result in different actions. For example, only the high level may activate the automatic braking system, whereas the collision alert system may be activated at both medium and high levels, for example. The threshold probabilities for different levels may be set as appropriate for the circumstances.

[0050] Fig. 5 schematically shows an example hardware configuration of the collision avoidance system. In this example, hardware components such as a sensing system 21 of the object detection system 2, a camera 31 and an accelerometer 32 of the trajectory determining system 3 may be operationally connected to a computer processor 10 configured to control the hardware components and perform processing on information received from the hardware components. The processor 10 may be configured to perform data processing functions of each of the object detecting system 2, the trajectory determining system 3 and the collision determining system 4.

[0051] The collision avoidance system 1 may be mounted to the vehicle. A plurality of sensors, receivers or cameras may be placed around the vehicle at required locations to capture the required information. It is possible that parts of the vehicle, such as boom arms, or equipment connected to the vehicle, such as trailers, may partially obstruct some sensors, receivers, or cameras. In such cases, the parts of the vehicle or equipment connected to the vehicle may be ignored by the object detecting system 2 and / or the collision detecting system 3. Alternatively, where possible, the sensors, receivers or cameras may be arranged so as not to capture information from the area in which the parts of the vehicle or equipment connected to the vehicle are located.

[0052] The collision avoidance system 1 is preferably configured to be retrofittable to the vehicle. Accordingly, the components of the object detecting system, trajectory determining system and collision determining system are preferably separate and distinct from the vehicle itself and configured to be connected to the vehicle.

[0053] It should be understood that the above described examples are illustrative only and variations are possible without departing from the spirit and scope of the application.

Claims

CLAIMS1. A collision avoidance system for an on-track plant vehicle, comprising: an object detecting system configured to detect an object within an area of interest surrounding the vehicle; a trajectory determining system configured to determine a trajectory of the vehicle; a collision determining system configured to determine a probability of collision occurrence based on an output of the object detection system and the trajectory determining system.

2. The collision avoidance system of claim 1, further comprising a collision alert system configured to generate an alert based on the determination by the collision determining system.

3. The collision avoidance system of claim 1 or 2, further comprising an automatic braking system configured to automatically slow the vehicle based on the determination by the collision determining system.

4. The collision avoidance system of any preceding claim, wherein the collision determining system is configured to assign an object detected by the object detecting system a probability of collision occurrence based on an intersection between the location of the detected object and the determined trajectory of the vehicle, a higher probability being assigned when there is intersection between the location of the detected object and the determined trajectory, and a lower probability being assigned when there is no intersection between the location of the detected object and the determined trajectory of the vehicle.

5. The collision avoidance system of any preceding claim, wherein the object detection system is configured to determine a trajectory of a detected object that is moving; and the collision determining system is configured to assign an object detected by the object detecting system a probability of collision occurrence based on an intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle, a higher probability being assigned when there is intersection between thedetermined trajectory of the detected object and the determined trajectory of the vehicle, and a lower probability being assigned when there is no intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle.

6. The collision avoidance system of claim 5, wherein a higher probability is assigned when a position of the object on its determined trajectory intersects a position of the vehicle on its determined trajectory at the same point in time, and a lower probability is assigned when a position of the object on its determined trajectory does not intersect a position of the vehicle on its determined trajectory at any point in time.

7. The collision avoidance system of any preceding claim, wherein the object detecting system is configured to determine whether a detected object is a known type of object; and the collision determining system is configured to assign a probability of collision occurrence based on the whether a detected object is determined to be a known type of object.

8. The collision avoidance system of any preceding claim, wherein the object detecting system comprises at least one transmitter configured to transmit a signal and at least one corresponding receiver configured to receive a signal reflected from an object.

9. The collision avoidance system of claim 8, wherein the at least one transmitter and at least one receiver comprises a Radar transmitter and receiver.

10. The collision avoidance system of claim 8 or 9, wherein the at least one transmitter and at least one receiver comprises a Lidar transmitter and receiver.

11. The collision avoidance system of any preceding claim, wherein the object detecting system comprises at least one video camera.

12. The collision avoidance system of any preceding claim, wherein the trajectory determining system comprises a rail detecting system configured to detect rails on which the vehicle travels and determine the path formed by the rails.

13. The collision avoidance system of claim 12, wherein the rail detecting system comprises at least one camera configured to image the rails.

14. The collision avoidance system of any preceding claim, wherein the trajectory determining system comprises a motion detecting system configured to determine parameters relating to the motion of the vehicle.

15. The collision avoidance system of claim 14, wherein the parameters include one or more of: direction of travel, speed, acceleration.

16. The collision avoidance system of claim 14 or 15, wherein the motion detecting system comprises an accelerometer.

17. The collision avoidance system of any preceding claim, wherein the motion detecting system comprises motion derived from tracking of object detections over time.

18. The collision avoidance system of any preceding claim, wherein the trajectory determining system comprises a location detecting system configured to determine a location of the vehicle.

19. The collision avoidance system of claim 18, wherein the location detecting system is a satellite positioning system, such as the Global Positioning System (GPS).

20. The collision avoidance system of any preceding claim, wherein the trajectory determining system comprises a route determining system configured to determine a route followed by the vehicle based on map information.

21. The collision avoidance system of any preceding claim, comprising one or more processors configured to perform data processing functions of each of the object detecting system, the trajectory determining system and the collision determining system.

22. A collision avoidance system for an on-track plant vehicle, comprising: an object detecting system configured to detect an object within an area of interest surrounding the vehicle;a trajectory determining system configured to determine a trajectory of the vehicle; a collision determining system configured to determine a probability of collision occurrence based on an output of the object detection system and the trajectory determining system; and a collision alert system configured to generate an alert based on the determination by the collision determining system; wherein: the collision determining system is configured to assign an object detected by the object detecting system a probability of collision occurrence based on an intersection between the location of the detected object and the determined trajectory of the vehicle, a higher probability being assigned when there is intersection between the location of the detected object and the determined trajectory, and a lower probability being assigned when there is no intersection between the location of the detected object and the determined trajectory of the vehicle, the object detection system is configured to determine a trajectory of a detected object that is moving; the collision determining system is configured to assign an object detected by the object detecting system a probability of collision occurrence based on an intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle, a higher probability being assigned when there is intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle, and a lower probability being assigned when there is no intersection between the determined trajectory of the detected object and the determined trajectory of the vehicle; the object detecting system comprises at least one Lidar transmitter configured to transmit a signal and at least one corresponding Lidar receiver configured to receive a signal reflected from an object; the trajectory determining system comprises a rail detecting system configured to detect rails on which the vehicle travels and determine the path formed by the rails, a motion detecting system configured to determine parameters relating to the motion of the vehicle, and a location detecting system configured to determine a location of the vehicle.

23. A method of collision avoidance for an on-track plant vehicle, comprising: detecting an object within an area of interest surrounding the vehicle; determining a trajectory of the vehicle;determining a probability of collision occurrence based on an output of the object detection and the trajectory determination.

Images