SYSTEM AND METHOD FOR DETECTING RAILWAY TRACK ANOMALIES

Abstract

 The present invention concerns a track anomaly detection system and a method for automatically detecting a surface anomaly of a railway track, the system comprising:
- a scanning system (11) configured for automatically scanning at least a portion of the track surface, said scanning system (11) comprising a laser emitter (111) and an array (112) of photodiodes (P), wherein the laser emitter is configured for emitting an incident ray towards the track surface according to a predefined angle (θ) and the array (112) of photodiodes (P) is configured for capturing light scattered by the track surface in response to the incident ray (R1);
- - a processing system (12) connected to the scanning system (11), said processing system (12) being con-figured for i) detecting an anomaly of the track surface by processing a signal outputted by each photodiode (P) in response to the captured light, ii) for associating to each detected anomaly a position, and iii) for signaling the detected anomaly by sending, for each detected anomaly, a signal configured for encoding said position of the detected anomaly.

Description

[0001] The present invention concerns a system and a method for detecting railway track anomalies.

[0002] The present invention is in particular directed to an automatic detection of an anomaly or defect or abnormal configuration (hereafter jointly referred to as "anomaly") of a track that can impact the safety of said track, the latter including at least one rail, sleepers, and fasteners. Railway track anomalies can jeopardize the safe motion of a guided vehicle moving on the rails of said track. By "Guided vehicle", it has to be understood any rail transport means running on at least one rail configured for supporting a wheel of said guided vehicle, said rail transport means being for instance public transport means like subways, trains or train units, etc., as well as load transporting means such as, for example, overhead traveling cranes, freight trains, for which safety is a very important factor and for which track defects may affect said safety and cause accidents.

[0003] By track anomaly it has to be understood any local divergence between a standard or conventional outside shape or surface of a track and a currently measured or image outside shape or surface. Such divergence can result for instance from a deterioration of an element of the track (e.g. of the rail itself, or of a fastener, or sleeper), or from the presence of an unusual object on the track, or from any other anomaly/defect whose presence can be detected by analyzing the track surface/shape and detecting a deviation of the analyzed track surface/shape from the standard/conventional shape/surface geometry of said track shape/surface. Such track anomalies comprise for instance rail deformation, broken elements like a broken sleeper, missing fastener, etc.

[0004] In order to detect such anomalies, i.e. divergences with respect to the standard surface or shape of a track, known in the art solutions have been proposed. Said solutions are notably based on :

• manual inspection of the track, typically involving a maintenance team regularly inspecting sections of track. This solution presents the disadvantage of being time consuming and prone to human error;
• image processing, wherein images of the track are automatically acquired and analyzed, e.g. using machine learning, for detecting any abnormal feature of the track. This solution requires unfortunately complex calculations and processing large amounts of data;
• Light Detecting and Ranging (LIDAR) technology, wherein an acquired 3D model of the track is compared to a standard 3D model for detecting divergences. This solution also requires the processing of large amounts of data.

[0005] The current solutions are thus complex to implement or prone to human errors, and simpler and efficient solutions are still needed.

[0006] An objective of the present invention is to propose a simple and efficient method and system for detecting railway track anomalies, notably capable of reducing the amount of data to be processed.

[0007] This objective is solved by the measures taken in accordance with the independent claims. Further advantageous embodiments are proposed by the dependent claims.

[0008] According to an aspect of the invention, a track anomaly detection system (hereafter TADS) is proposed for automatically detecting a surface anomaly of a railway track of a railway network, said TADS being configured for being installed on a guided vehicle, and comprising:

• a scanning system configured for automatically scanning at least a portion of the track surface, said scanning system comprising a laser emitter and an array of photodiodes, wherein the laser emitter is configured for emitting an incident ray towards the track surface according to a predefined angle and the array of photodiodes is configured for capturing light scattered by the track surface in response to the incident ray. Preferentially, the array of photodiodes is centered, notably according to its length, on a specular reflection of the incident ray when the latter would hit a reflecting surface (i.e. a mirror-like surface) when a plane defined by the light absorption surface of the array of photodiodes is parallel to the reflecting surface (which would represent the track for instance);
• a processing system connected to the scanning system, said processing system being configured for identifying or detecting an anomaly of the track surface by processing a signal (e.g. current signal) outputted by each photodiode of the array in response to, or resulting from, said capture of light by the concerned photodiode. For instance, the processing system might be configured for identifying or detecting said anomaly from a variation, notably a temporal variation, of an intensity value of said signal outputted by each photodiode of the array. The processing system is further configured for associating to each detected anomaly a position, e.g. a geographic position, enabling to locate the detected anomaly within the railway network. The processing system is finally configured for signaling the detected anomaly by sending, for each detected anomaly, a signal configured for indicating the position of said anomaly. Said signal might be sent to an operator, or to a control center, or to a device in charge of the management of railway network alerts. Optionally or alternatively, the position of each detected anomaly might be stored by the processing system in a database or memory.

[0009] Preferentially, the TADS further comprises an imaging system configured for being installed on-board the guided vehicle, backwards the scanning system when considering a forward motion of the guided vehicle, and configured for acquiring an image of at least a part of said track surface portion scanned by the scanning system, wherein said part comprises the detected anomaly. In order to acquire said image of said part of scanned track surface that comprises the detected anomaly, the imaging system is connected to the processing system and the latter is configured for triggering the time at which the acquisition of the image takes place after each detection of an anomaly. Then, for each detected anomaly, the acquired image of the anomaly and its position might be provided via an interface or encoded in said signal configured for signaling the anomaly. Optionally or alternatively, for each detected anomaly, said position and image might be stored in said database or memory. Optionally, the processing system might be configured for processing the acquired images, performing for instance automatic anomaly recognition or identification using known in the art techniques, e.g. based on deep learning. Preferably, instead of sending said signal for each detected anomaly, then said signal is only sent for detected anomalies that have been identified or recognized as belonging to a predefined set of anomaly types (e.g. "obstacle" when detecting the presence of an unusual object on the track surface, or "fastener" when detecting a missing or broken fastener, or "rail defect" when detecting a defect of the running surface of the rail such as a crack, shelling, wear, etc. ), said signal encoding for instance the type of anomaly, its position, and its image. Optionally, said signal may also automatically trigger a sending of an alarm via said interface, e.g. to a control center or to an operator or to a device in charge of managing alarms for the railway network.

[0010] According to the present invention, an anomaly is thus detected by analyzing the signal outputted by each photodiode of the array, when considering notably its position within the array. In particular, different patterns of signal intensities outputted by the photodiodes of the array might be predefined or stored in a database as reference patterns, each for a different type of track surface or geometry or track element. For instance, a first reference pattern might be stored in said database for enabling a detection of an anomaly of the rail, another reference pattern for the detection of an anomaly of a first type of sleepers, another for the detection of an anomaly of a second type of sleepers, another one for the anomaly of the fasteners, etc. Each reference pattern provides or defines for instance for each photodiode of the array (i.e. in function of the position of the photodiode within the array) a reference signal intensity value to be used by the processing system for the anomaly detection e.g. by comparison with currently acquired signal intensity values. Said reference signal intensity value might be a function of the time and/or defined for a given speed of the guided vehicle and/or in function of a type of track surface and/or in function of a scanned track element and/or in function of a path followed by the incident ray on the track surface. This enables the processing system to determine, during motion of the guided vehicle, whether the signals outputted by the photodiodes represent standard values (i.e. can be considered as equivalent to the reference values) or whether some measured values indicate a presence of an anomaly. For instance, if a pattern defines, for a given track surface, that a set of photodiodes of the array shall be each characterized by an output signal intensity value constantly above a threshold value while all other photodiodes are characterized by output signal intensity values constantly below said threshold, then the processing unit is able to detect an anomaly as soon as one or several of said other photodiodes output a signal characterized by an intensity value above said threshold. This provides a very simple and efficient way to detect an anomaly.

[0011] Additionally and in particular, the present invention proposes to implement a discontinuous imaging of the track, wherein said imaging occurs only temporary, i.e. during a predefined finite period of time, at or after each detection of an anomaly by the processing system. In other words, at or after each detection, the imaging system is launched by the processing system for a predefined period of time and then automatically stopped by the processing system until the next anomaly is detected. The starting and stopping of the acquisition of one or several images is preferably controlled by the processing system in function of the speed of the guided vehicle and/or in function of the distance separating the scanning system from the imaging system so that the time at which the image acquisition takes place corresponds to the time at which the detected anomaly is in the field of view of the imaging system. The detection of anomalies by the processing system is thus configured for triggering an acquisition of one or several images of a track surface part comprising the scanned portion of track surface, said acquisition being discontinuous, i.e. temporally limited, starting after a first predefined time interval after each detection, and lasting for a second predefined time interval, wherein said first and second time interval preferentially vary in function of the speed of the guided vehicle for a given fixed distance separating the scanning system from the imaging system.

[0012] According to another aspect of the present invention, a method for automatically detecting a surface anomaly of a railway track by a TADS installed on-board a guided vehicle moving on said track is proposed. Said method comprises the following steps:

• automatically scanning at least a portion of track surface by means of a scanning system, wherein said scanning comprises emitting towards the track surface an incident ray of light and capturing by an array of photodiodes and in response to the emission of said incident ray light scattered by the track surface;
• detecting an anomaly of said track surface by processing by means of a processing system a signal outputted by each photodiode of said array. Said processing might comprise comparing a pattern of signals outputted by the array of photodiodes to a reference pattern;
• for each detection of an anomaly, automatically determining a position at which said anomaly has been detected, and optionally, automatically triggering an acquisition of an image of a part of the scanned portion of track surface, said part comprising the detected anomaly. In other words, the processing system is configured for automatically triggering an acquisition of an image of the detected anomaly by the imaging system;
• signaling the detected anomaly by sending, for each detected anomaly, a signal configured for indicating the position of said anomaly. Optionally, said signal is configured for encoding said acquired image of the anomaly and its position. Preferentially, said position, and said image of the anomaly if acquired, are stored in a database or memory.

[0013] Further aspects of the present invention will be better understood through the following drawings, wherein like numbers designate like objects:

Figure 1

schematic lateral view of a preferred embodiment of a TADS according to the invention installed onboard a guided vehicle;

Figure 2

flowchart of a preferred method according to the invention;

Figure 3

schematic front view of a preferred embodiment of TADS according to the invention installed on-board a guided vehicle;

Figure 4

details of a scanning system according to the invention;

Figure 5

schematic representation of a first scanning path of a scanning system according to the invention;

Figure 6

schematic representation of a second scanning path of a scanning system according to the invention;

Figure 7

schematic examples of signal intensities outputted by each photodiode of an array according to the invention.

[0014] Figure 1 shows a preferred embodiment of a TADS according to the invention, which is configured for detecting anomalies of a surface of the track 2. Said TADS is configured for being installed on board a guided vehicle 1, for instance a train or a metro, and comprises a scanning system 11 configured for scanning the track surface facing the underside of the guided vehicle 1, a processing system 12 for processing data or signals resulting from said scanning, and optionally an imaging system 13 for acquiring one or several images of a part of the portion of scanned track surface. The imaging system 13 is located beyond (or downstream) the scanning system 11 with respect to a travel direction F of the guided vehicle 1 and comprises one or several cameras configured for imaging said portion of track that comprises the track surface scanned by the scanning system 11. This makes it possible to trigger by the processing system 12 sporadic acquisition of images by the imaging system 13 for each detection, by the processing system 12, of an anomaly in said signal or data outputted by the scanning system 11.

[0015] By surface of the track 2, it has to be understood the surface of the elements forming the track 2, such as the rail 22, fasteners 23, sleepers 21 (or a track ground on which rails are fixed) as illustrated in Fig. 3, and which are facing the scanning system when the latter moves over the track 2 during a travel of the guided vehicle on the railway network. For instance, the rail 21 has usually a well-defined structure comprising typically a foot, a web and a head that serves as running surface for wheels of the guided vehicle.

[0016] The TADS according to the invention might be installed on the guided vehicle 1 and configured for scanning only said running surface of the rail during the motion of the guided vehicle over said rail. According to other embodiments, the TADS may comprise several scanning systems 11, each dedicated to the scanning of a specific element of the track 2: e.g. a first scanning system is configured for scanning said running surface only, and/or a second scanning system is configured for scanning fasteners, and/or a third scanning system is configured for scanning a part of the track comprised between the pair of rails 22 or between the fasteners, etc. Any surface of the track 2 that is facing the scanning system 11 can thus be scanned by one or several scanning systems 11 according to the invention, and data or signals resulting from this scanning subsequently analyzed by the processing system 12 in order to detect anomalies of the track 2 and optionally trigger an acquisition of an image of said part of scanned track surface comprising the detected anomaly by the imaging system 13. For instance, for each scanning system 11 scanning a portion of track surface, at least one camera of the imaging system is configured for acquiring an image of the scanned track surface portion. In other words, the imaging system might comprise for instance a first camera for acquiring an image of the running surface scanned by the first scanning system, a second camera for acquiring an image of the fasteners scanned by the second scanning system, and a third camera for acquiring an image of the portion of track surface comprised between the pair of rails 22 or between the fasteners and that is scanned by the third scanning system. The signals outputted by the array of each scanning system are then processed by said processing system 12.

[0017] The method according to the invention will now be described in more details with respect to Fig. 2, together with Fig. 1, 3-6:
At step 201, the scanning system 11 is configured for automatically scanning at least a portion of the track surface. The scanning path realized by the scanning system 11 may extend in two dimensions as shown by the first scanning path P1 or the second scanning path P2 in Fig. 5, or in one dimension, as shown by the third scanning path P3 in Fig. 6. A scanning path in one dimension means that the path followed by the incident ray on the track surface, e.g. on the running surface of the rail, is a straight line extending in a single direction. A scanning path in two dimensions as shown in Fig. 5 means that the path followed by the incident ray on the track surface extends in two dimensions (2D), being for instance a series of interconnected straight segments, wherein two adjacent segments are connected to each other according to a predefined non-zero angle, or usually a plane curve, or any other line extending in at least two dimensions of a plane. The geometry of the 2D scanning path will depend of course on the speed of the guided vehicle and as explained afterwards, on the motion of the laser emitter 111, notably its rotational speed around a rotational axis A.

[0018] Indeed, and as shown in Fig. 4, the scanning system 11 comprises a laser emitter 111 and an array of photodiodes 112. The laser emitter 111 is configured for emitting an incident ray R1 towards the track surface, hitting the latter at a position S1 according to a predefined fixed angle θ. This incident ray R1 is then scattered by the track surface at many different angles, except, of course, if said track surface is a mirror-like surface, in which case there would be, theoretically, only pure specular reflection (represented by the ray R2). Usually, the light scattered by the track surface is thus a combination of specular and diffuse reflection. The array 112 of photodiodes is configured and arranged for capturing said scattered light, each photodiode outputting then a signal resulting or that is a function of the amount of light captured. In particular, said array 112 comprises multiple photodiodes P arranged longitudinally one after another, i.e. aligned with one another, extending thus longitudinally according to a length L. A set P1 of central photodiodes P, i.e. one or several photodiodes P located in the middle of the array with respect to its length (i.e. at approximately L/2), is preferentially configured for capturing a reflected ray R2 that would result from the incident ray R1 hitting a mirror-like surface. In other words, if for instance the laser source of the laser emitter is positioned within the same plane as the plane comprising the light absorption surface of the photodiodes (hereafter absorption plane), then the distance D_s separating the center of said set P1 (wherein "center" is defined as half the longitudinal length of said set of photodiodes) from the laser source is approximately equal to D_s = 2·D/tanθ, wherein D is the distance from said absorption plane to the scanned track surface 2 (considered as constant in the present case) and θ the angle defined between the incident ray R1 and said absorption plane. Preferentially, the laser emitter 111 and the array of photodiodes P are arranged so that the absorption plane is parallel to the track surface, the incident ray R1 and the reflected ray R2 are in a plane perpendicular to the absorption plane.

[0019] Preferentially, the laser emitter 111 is aligned with the array 112, i.e. it is located within a longitudinal extension of the array. In particular, the array 112 and the laser 111 are supported by a same supporting structure. According to a preferred embodiment, said supporting structure is configured for being driven into rotation W around an axis of rotation A by a motor 113 configured for being mounted on the underside of the guided vehicle, e.g. fixed to the body or housing or bogie of the guided vehicle. In this case, the position of the laser emitter 111 on said supporting structure and the angle θ are arranged and defined for having the position S1 different from an intersection position, wherein said intersection position is defined as the intersection of an extension of the rotation axis A with the track surface. This preferred embodiment makes it possible to create 2D scanning paths along the track surface as shown in Fig. 5 by rotating said supporting structure, and thus the laser emitter 111 and the array 112, around the rotation axis A. Optionally, an additional laser emitter might be mounted on said supporting structure. In such a case, one laser might be installed at one extremity of the array 112, and the additional laser at the other extremity of said array 112, each aligned with the array 112. If the scanning system 11 is free of said motor 113, or if the motor 113 is stopped, then straight scanning paths, as illustrated by Fig. 6, can be implemented by the scanning system 11. Preferably, the processing system 12 is configured for controlling the rotational speed of the supporting structure around the axis A, i.e. the motor, notably in function of the speed of the guided vehicle 1, and in particular so that a 2D path comprising scanning lines sufficiently close to one another (e.g. the distance D_L measured parallel to the rail, passing through said intersection point, and separating two consecutive lines of said path is between 1-5mm - see Fig. 5) is obtained, so that a "continuous-like" surface S be scanned (i.e. when considering a track surface, the density of points of said surface hit by the incident ray R1 is sufficiently high for considering the whole surface as scanned and not simply a line). Such surface S would correspond for instance to the area enclosed within the dotted lines in Fig. 5.

[0020] During the scanning of the track surface, the photodiodes P capture light scattered by the track surface and output each a signal (e.g. current) whose intensity is proportional to the intensity of the captured light. Figure 7 show two different graphs, G1 and G2, which represent two different patterns of the signal intensities outputted by the photodiodes of the array. Indeed, the amount of light captured by each photodiode, and consequently the intensity of the signal outputted by each photodiode, depend on the type and geometry of track surface and on the longitudinal position of the considered photodiode within the array. For sufficiently homogeneous surfaces, the set P1 of photodiodes will usually output signals with the highest intensity values, since said set P1 will capture specular and diffuse reflection, while the other photodiodes will mainly capture only diffuse reflection.

[0021] The processing system 12 acquires then the signal outputted by each photodiode of the array and, at step 202, detects an anomaly of the track surface by processing the acquired signals. The processing system 12 typically comprises a processor and a memory. Different methods might be implemented by the processing system 12 for detecting an anomaly. The chosen method will depend notably on the type and geometry of track surface that is scanned by the scanning system 11. For instance, for homogeneous flat surfaces, such as the running surface of the rail, the processing system might be configured for automatically determining whether signal intensities greater than a predefined threshold T1 or T2 belong only to a predefined set of photodiodes. If this is the case, then the processing system indicates no anomaly, i.e. no signal is sent. If it is not the case, then the processing system indicates or signals an anomaly.

[0022] According to other detection methods, the processing system 12 might be configured for comparing the signal intensities outputted by the photodiodes of the array to reference patterns of signal intensities, wherein each reference pattern might be stored in a database of the TADS and is configured for defining for each photodiode of the array a reference intensity value in function of a scanning path and the track surface on which said scanning path takes place. Said reference pattern might define for each diode temporal variations of the intensity values outputted by each diode for said scanning path on said track surface, wherein said intensity values are used by the processing system as references when determining whether current intensity values outputted by the photodiodes of the array are "normal" (i.e. the absolute value of the difference between a current value acquired by the processing system from a current scanning of the track surface and a reference value is smaller or equal to a predefined value) or indicate the presence of an anomaly (i.e. if said absolute value of the difference is greater than the predefined value). Indeed, the processing system might be configured for correlating currently acquired signal intensities values outputted by the photodiodes of the array with intensity values defined or associated to a reference pattern, by matching patterns, comparing values, or otherwise processing said values in order to determine any divergence between the reference pattern and the currently acquired intensity values, i.e. a current pattern of intensity values outputted by the array.

[0023] At step 203, the processing system 12 automatically determines, for each detected anomaly, a position of said anomaly. For this purpose, it can use any known in the art positioning system, like GPS systems, or acquire said position from a positioning system of the guided vehicle. The position according to the invention is configured for enabling locating the anomaly with respect to the railway network. For instance, it preferentially provides coordinates, e.g. geographic coordinates, of the anomaly that enable to locate said anomaly within the railway network. Typically, the position of an anomaly is assimilated or considered as equal to the position of the scanning system at the time the photodiode output signal which led to the detection of said anomaly has been outputted, the position of the scanning system being provided by any known in the art positioning system, like a GPS system.

[0024] At step 204, the processing system 12 is configured for signaling the detected anomaly by sending, for each detected anomaly, a signal configured for indicating or encoding the position of said anomaly. Said signal might be received for instance by an operator or a device in charge of the management of alerts for the railway network.

[0025] Optionally, for each detected anomaly, the processing system 12 automatically triggers an acquisition of an image of said anomaly by the imaging system 13. For this purpose, the processing system 12 might be configured for automatically controlling a start of an image acquisition by the imaging system 13 at or after each detected anomaly. Said image acquisition may start at the time at which the anomaly is detected by the processing system, or after said detection. The time at which the image acquisition takes place is controlled by the processing system 12 in order to enable the acquisition of one or several images of a part of the portion of the scanned track surface, wherein at least one of the images comprises the detected anomaly. In particular, if a detection takes place at the time T0, then the time T at which the image acquisition is performed is a function of the speed V of the guided vehicle and a fixed distance d separating the scanning system from the imaging system: T = T0 + d/V - t', wherein t' is a positive value accounting for the reactivity of the imaging system when receiving a command of image acquisition from the processing system. In particular, the processing system might be connected to a control system of the guided vehicle in order to acquire real time information regarding the speed or position in function of the time of the guided vehicle, and using said information for triggering the acquisition of said image of the scanned track surface so that the acquired image of said scanned track surface comprises the detected anomaly. Then, for each detected anomaly, said image and the position of the anomaly might be encoded in said signal and provided to an operator or control center or device in charge of managing railway network alerts. Preferentially, the image acquisition automatically stops after a predefined time, or the processing system automatically stops said acquisition after a predefined time, which can depend on the guided vehicle speed. This provides the advantage of decreasing the volume of data related to image acquisition, since the latter takes place discontinuously, only at each detection of anomaly, and for a predetermined duration.

[0026] To conclude, the present invention proposes a simple system and efficient system capable of detecting anomalies of a track surface by analyzing signals outputted by an array of photodiodes, and optionally acquiring images of the detected anomaly, and thus of the track surface, only in case of detection of an anomaly, avoiding thus to image and process images of complete sections of track.

Claims

1. Track Anomaly Detection System - hereafter TADS - (1) for automatically detecting a surface anomaly of a railway track (2), said TADS being configured for an installation on-board a guided vehicle (1), and comprising:

- a scanning system (11) configured for automatically scanning at least a portion of the track surface, said scanning system (11) comprising a laser emitter (111) and an array (112) of photodiodes (P), wherein the laser emitter (111) is configured for emitting an incident ray (R1) towards the track surface according to a predefined angle (θ) and the array (112) of photodiodes (P) is configured for capturing light scattered by the track surface in response to the incident ray (R1);

- a processing system (12) connected to the scanning system (11), said processing system (12) being configured for i) detecting an anomaly of the track surface by processing a signal outputted by each photodiode (P) in response to the captured light, ii) for associating to each detected anomaly a position, and iii) for signaling the detected anomaly by sending, for each detected anomaly, a signal configured for encoding said position of the detected anomaly.

2. TADS according to claim 1, comprising an imaging system (13) configured for being installed backwards the scanning system (11) when considering a forward motion of the guided vehicle (1), said imaging system (13) being configured for imaging at least a part of the track surface comprising the track surface scanned by the scanning system (11).

3. TADS according to claim 2, wherein the processing system (12) is configured for triggering an acquisition of an image of said part of the track surface for each detected anomaly.

4. TADS according to claim 3, wherein said triggering is a function of the speed of the guided vehicle (1) and the distance separating the imaging system (13) from the scanning system (11).

5. TADS according to one of the claims 2-4, wherein said imaging of at least a part of the track surface is temporally discontinuous.

6. TADS according to one of the claims 1-5, wherein the processing system (12) uses reference patterns for detecting track surface anomalies, wherein each reference pattern defines, for each photodiode (P) of the array (112), a reference signal intensity value.

7. TADS according to one of the claims 1 to 6, wherein the laser emitter (111) is located within a longitudinal extension of the array (112).

8. TADS according to one of the claims 1 to 7, wherein the laser emitter (111) and the array (112) are mounted on a same supporting structure.

9. TADS according to claim 8, wherein said same supporting structure is configured for being driven into rotation by a motor (113).

10. TADS according to claim 9, wherein the processing system (12) is configured for controlling said motor (113) in order to control a rotational speed of the supporting structure.

11. Method for automatically detecting a surface anomaly of a railway track by a track anomaly detection system, hereafter "TADS" installed on-board a guided vehicle (1) configured for moving on said railway track (2), the method comprising:

- automatically scanning (201) at least a portion of a surface of the railway track (2) by means of a scanning system (11), wherein said scanning comprises emitting an incident ray (R1) towards the track surface and capturing by an array (112) of photodiodes (P) and in response to the emission of said incident ray (R1) light scattered by the track surface;

- detecting (202) an anomaly of said track surface by processing by means of a processing system (11) a signal outputted by each of the photodiodes (P) of said array (112) in response to the captured light;

- automatically determining (203), for each detected anomaly, a position of the anomaly;

- signaling (204) each detected anomaly by sending a signal configured for encoding the position of the detected anomaly.

12. Method according to claim 11, wherein processing said signal comprises comparing the signals outputted by the array of photodiodes to a reference pattern.

13. Method according to claim 11 or 12, comprising triggering an acquisition of an image of each detected anomaly by an imaging system (13).

14. Method according to one of the claims 11 to 13, comprising driving into rotation a supporting structure configured for supporting a laser emitter (111) and the array (112) .

15. Method according to claim 14, comprising controlling the rotational speed of the supporting structure by the processing system (11).

Drawing