AUTOMATIC REAL-TIME DATA GENERATION

Abstract

The present invention discloses a system and a method for automatic real-time data generation, particularly in a railway infrastructure. This is facilitated by providing a processing component, a model analyzer, wherein the model analyzer is configured to generate at least one simulation model and a weight analyzer. The weight analyzer is configured to associated statistical weight to at least on infrastructural feature.

Claims

1. A system, comprising: a. at least one processing component; b. at least one storage component; c. a plurality of sensor nodes; wherein, the processing component is configured to receive sensor data from the at least one sensor node, d. at least one model analyzer, configured to generate at least one simulation model; and e. a weight analyzer, configured to automatically associate a statistical weight to at least one infrastructural feature.

2. The system according to claim 1 wherein the weight analyzer is configured with machine learning techniques, such as deep learning techniques, further configured to self-leam at least one infrastructural feature.

3. The system according to claim 1 wherein the sensor node comprises a pressure sensor and/or accelerometer and/or inclinometer and/or thermal sensor and/or acoustic sensor and/or strain gauge sensor and/or water pressure sensor and/or linear variable displacement sensor and/or visual sensor and/or load sensors and/or any combination thereof.

4. The system according to claim 1 wherein the model analyzer is further configured to automatically generate the at least one infrastructural feature using the simulation model.

5. The system according to claim 1 wherein the processing component is configured to generate a database using the sensor data.

6. The system according to claim 1 wherein the processing component is further configured to automatically classify at least a portion of the database, using the machine learning techniques, such as pattern recognition.

7. The system according to claim 1 wherein the processing component is further configured to store the database and/or classified database in the storage component.

8. The system according to claim 1 wherein the model analyzer is configured to generate the simulation model based on finite element method and/or multi-body simulation method and/or finite difference method and/or lumped parameter method and/or any combination thereof.

9. The system according to claim 1 wherein the system further comprises a noise decoder, the noise decoder comprises machine learning techniques, and is configured to automatically determine a noise pattern in the database.

10. The system according to claim 1 wherein the model analyzer is configured to generate at least one noise model, based on the database, and further configured to fuse the noise model to the simulation model.

11. The system according to claim 1 wherein the model analyzer is further configured to generate at least a portion of the simulation model based on a user input.

12. A method, comprising the steps of: a. obtaining sensor data from at least one or a plurality of sensor node(s); b. generating simulation model(s); c. automatically fusing at least portion of the sensor data with the simulation model; and d. automatically predicting at least one infrastructural feature, preferably associated with the sensor node.

13. The method according to claim 12 wherein the method comprises the step of carrying out the method on the system according to claim 1.

14. A device comprising: a. a device processing component, configured for an interactive model analysis; b. an interface, configured to pull user input; and c. a memory component, configured to store the user input.

15. A computer program product comprising instructions, when the program is executed by any of the system claims causes the system to perform the method steps according to any of the method claims.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0201] The present invention will now be described with reference to the accompanying drawings, which illustrate embodiments of the invention. These embodiments should only exemplify, but not limit, the present invention.

[0202] FIG. 1 schematically depicts an embodiment of a sensor node routing in a railway infrastructure.

[0203] FIG. 2 depicts a system embodiment according to an aspect of the present invention.

[0204] FIG. 3 schematically illustrates a data flow diagram related to the system.

[0205] FIG. 4 schematically shows an exemplary operation of the system.

[0206] FIG. 5 depicts the steps of a method according to an aspect of the present invention.

[0207] FIG. 6 depicts an exemplary representation of sensor data, particularly load data.

DETAILED DESCRIPTION OF THE DRAWINGS

[0208] In the following description, a series of features and/or steps are described. The skilled person will appreciate that unless explicitly required and/or unless requires by the context, the order of features and steps is not critical for the resulting configuration and its effect. Further, it will be apparent to the skilled person that irrespective of the order of features and steps, the presence or absence of time delay between steps can be present between some or all of the described steps.

[0209] It is noted that not all the drawings carry all the reference signs. Instead, in some of the drawings, some of the reference signs have been omitted for sake of brevity and simplicity of illustration. Embodiments of the present invention will now be described with reference to the accompanying drawings.

[0210] FIG. 1 illustrates an embodiment of a sensor node 1-9 routing in a railway infrastructure. There is shown an example of a railway section with the railway itself, comprising rails and sleepers. Instead of the sleepers also a solid bed for the rails can be provided. Moreover, a mast that is just one further example of constructional elements that are usually arranged at or in the vicinity of railways. A sensor node 1-9 can be arranged on one or more of the sleepers. The sensor 10 can comprise an acceleration sensor and/or any other kind of railway specific sensor. The sensor node 1-9 can further comprise a wireless sensor network. The sensor node can transmit data to a base station (not shown here). The base station can be installed to the railway infrastructure. The base station can also be installed in the surroundings of the railway infrastructure. The base station can also be a remote base station. The communication module between the base station and the sensor node (s) can comprise, for example Xbee with a frequency of 868 MHz.

[0211] The sensor node(s) 1-9 can also be installed in cases and inserted inside the railway infrastructure, for example inside a special hole carved into the concrete. The case can also be attached to the railway infrastructure using fixers. The sensor node 1-9 can be obtaining sensor data based on acceleration, inclination, distance, etc.

[0212] The sensor node 1-9 may further be divided into group, for example based on the distance. The sensor node 1-9 lying within a pre-determined distance may be controlled by one base station. The sensor node 1-9 can also be installed on the moving railway infrastructure such as on-board of a vehicle. The sensor node 1-9 can comprise an amplifier to amplify any signal received by the base station.

[0213] The sensor nodes 1-9 can be installed such that the sensor node lying within one group can communicate with their base station in one-hop. The base station can receive information from its ‘neighbors’ and retransmit all the information to the server 800.

[0214] The sensor node 1-9 can comprise sensor(s). The sensor can be accelerometers, such as Sensor4PRI for example ADCL 345, SQ-SVS etc. The sensor node 1-9 can comprise inclinometers, such as SQ-SI-360DA, SCA100T-D2, ADXL345 etc.

[0215] The sensor node can further comprise distance sensors. The distance sensors can be configured to at least measure the distance between slab tracks, using infrared and/or ultrasonic. The distance sensor can be for example, MB1043, SRF08, PING, etc.

[0216] The sensor node 1-9 can comprise visual sensors, such as 3D cameras, speed enforcement cameras, traffic enforcement cameras, etc. It may be noted that sensor node 1-9 may comprise sensors to observe the physical environment of the infrastructure the sensor node 1-9 are installed in. For example, temperature sensor, humidity sensor, altitude sensor, pressure sensor, GPS sensor, water pressure sensor, piezometer, multidepth deflectometers (MDD), etc.

[0217] The sensor node 1-9 can be installed to the railway structure depending on the sensor. For example, the strain gauge sensor can be most efficient when installed to the rail. The piezometer can be installed to the sub-ballast. The LVDT sensor can be installed to the sleeper. One sensor node 1-9 can be installed to more than one places.

[0218] The sensor node 1-9 can be installed according to a protocol based on routing trees to be able to transmit information to the base station. Once the information has been received, the UMTS technology can be used to send sensor data to a remote server 800. The sensor node 1-9 can comprise an analog-to-digital converter, a micro controller, a transceiver, power and memory. One or more sensor(s) can be embedded in different elements and can be mounted on boards to be attached to the railway infrastructure. The sensor node 1-9 can also comprise materializing strain gauges, displacement transducers, accelerometers, inclinometers, acoustic emission, thermal detectors, among others. The analog signal outputs generated by the sensors can be converted to digital signals that can be processed by digital electronics. The data can then be transmitted to the base station by a microcontroller through a radio transceiver. All devices can be electric or electronic components supported by power supply, which can be provided through batteries or by local energy generation (such as solar panels), the latter mandatory at locations far away from energy supplies.

[0219] The sensor data 101 collected from the sensor nodes 1-9 can be transferred to the base station using wireless communication technology such as CAN, FlexRay, Wi-Fi or Bluetooth. For example, the ZigBee network can be advantageous to consumes less power. On the other hand, for transmitting the input 101 data from the base station to the server 800 long-range communication such as GPRS, EDGE, UMTS, LTE or satellite can be used. Due to the short transmission range, communications from sensor nodes may not reach the base station, a problem that can be overcome by adopting relay nodes to pass the data from the sensor nodes 1-9.

[0220] FIG. 2 depicts a system according to an aspect of the present invention. The server 800, The collected sensor data 101 can be transmitted to the server 800 server through long-range communications such as GPRS, EDGE, UMTS, LTE or satellite. The sensor node 1-9 can also communicate directly with the server 800 without requiring the use of the base station as a gateway.

[0221] The server 800 may comprise a data transmitting component may be configured to establish a bidirectional communication with the base station. In other words, the server 800 may retrieve sensor data 101 from the base station, and further may provide it to the processing component 100, for example, vibrational data.

[0222] In one embodiment, the server 800 may comprise a cloud server, a remote server and/or a collection of different type of servers. Therefore, the server 800 may also be referred to as cloud server 800, remote server 800, or simple as servers 500. In another embodiment, the servers 800 may also converge in a central server.

[0223] It will be understood that the server 800 may also be in bidirectional communication with a storage component and an interface component. The storage component may be configured to receive information from the server 800 for storage. In simple words, the storing component 800 may store information provided by the servers 800. The information provided by the server 800 may include, for example, but not limited to, data obtained by sensor nodes 1-9, data processed by the processing component 100 and any additional data generated in the servers 800 or the processing component 800.

[0224] It will be understood that the servers 800 may be granted access to the storage component comprising, inter alia, the following dictions about future or otherwise unknown events.

[0225] The storage component can comprise comprises a volatile or non-volatile memory, such as a random-access memory (RAM), Dynamic RAM (DRAM), Synchronous Dynamic RAM (SDRAM), static RAM (SRAM), Flash Memory, Magneto-resistive RAM (MRAM), Ferroelectric RAM (F-RAM), or Parameter RAM (P-RAM).

[0226] It will also be understood that the term server may also refer to a computer program, and/or a device, and/or a plurality of each or both that may provide functionality for other programs, devices and/or components of the present invention. For instance, a server may provide various functionalities, which may be referred to as services, such as, for example, sharing data or resources among multiple clients, or performing computation and/or storage functions. It will further be understood that a single server may serve multiple clients, and a single client may use multiple servers. Furthermore, a client process may run on the same device or may connect over a network to a server on a different device, such as a remote server or a cloud. The server may have rather primitive functions, such as just transmitting rather short information to another level of infrastructure, or can have a more sophisticated structure, such as a storing, processing and transmitting unit.

[0227] The processing component 100 can comprise a CPU (central processing unit), GPU graphical processing unit), DSP (digital signal processor), APU (accelerator processing unit), ASIC (application-specific integrated circuit), ASIP (application-specific instruction-set processor) or FPGA (field programmable gate array) or any combination thereof.

[0228] The processing component 100 can further be generating the structured database 103 using the sensor data 101. The structured database 103 may comprise. The processing component 100 can be configured to automatically recognize the sensor associated with the sensor data 101 and can further generate structured database 103 based on the type of the sensor.

[0229] The processing component 100 can be configured with machine learning techniques, such as pattern recognition. The processing component can further be configured to generate labeled data using the structured database 103 and/or the sensor data 101.

[0230] The processed data, meaning the data transmitting from the processing component 100 which can comprise the structured database and/or the labeled data. The processed data can be then automatically pulled by the model analyzer 300. The model analyzer 300 can comprise generating at least one simulation model 102 based on at least the physical conditions (temperature, waves, speed, etc.).

[0231] The model analyzer 300 may comprise of a computer program product which can be configured to be programmed based on at least one of dynamical systems, statistical models, differential equations, game theoretic models, logic. The model analyzer 300 can be equipped with neural networks. The model analyzer 300 can further be configured to automatically learn the at least one of governing equations, assumptions, constraints using an existing knowledgebase. The model analyzer 300 can also learn using the sensor data 101 and/or structured database 103.

[0232] The model analyzer 300 can also be configured to generate at least one noise model 104 based on at least one of dynamical systems, statistical models, differential equations, game theoretic models, logic.

[0233] The simulation model 102 and/or the noise model 104 generated by the model analyzer 300 can be automatically fed to the weight analyzer 501/500. The weight analyzer 501/500 can comprise a machine learning classifier. The weight analyzer 500/501 may be trained using the simulation model 102 to generate labeled data. The weight analyzer 500/501 can be configured to generate the labeled data by using at least one of k-nearest neighbor, case-based reasoning, artificial neural networks, Naïve Bayes, etc.

[0234] The weight analyzer 501/500 can further be configured to predict at least one infrastructural feature (ballast, frog, geometry, speed, etc.) based on the labeled data and can further transmit the results to a user device 200.

[0235] The user device 200 can comprise a memory component such as, main memory (e.g. RAM), cache memory (e.g. SRAM) and/or secondary memory (e.g. HDD, SDD). The user device 200 may also comprise at least of an output user interface, such as: screens or monitors configured to display visual data (e.g. displaying graphical user interfaces of the questionnaire to the user), speakers configured to communicate audio data (e.g. playing audio data to the user). The user device 200 can also comprise an input user interface, such as, camera configured to capture visual data (e.g. capturing images and/or videos of the user), microphone configured to capture audio data (e.g. recording audio from the user), and a keyboard configured to allow the insertion of text and/or other keyboard commands (e.g. allowing the user to enter text data and/or another keyboard and mouse, touchscreen, joystick—configured to facilitate the navigation through different graphical user interfaces of the questionnaire.

[0236] FIG. 3 depicts an embodiment according to the present invention. The figure particularly represents the training phase of the weight analyzer 500. The weight analyzer 500 can comprise ‘self-learning’ of the features. The self-learning of the weight analyzer 500 can be by using the simulation model 102, sensor data 101, structured database 103, noise model 104, etc. The weight analyzer 500 can comprise building at least one machine learning model based on the simulation model 102 and then further associate statistical weights to the at least one feature.

[0237] FIG. 4 depicts a deployment phase of the weight analyzer 501, after it is trained. The weight analyzer 501 can be pulling in the sensor data 101 directly from server 800. The weight analyzer 501 may be configured to generate at least one feature prediction and transmit it to the user device 200.

[0238] FIG. 5 depicts a method to train the weight analyzer 501/500 using fusion of the noise model 104. The model analyzer 300 can comprise generative machine learning techniques such as an autoencoder to generate at least one noise model 104. The noise model 104 can further be fused with the simulation model 102 to generate the realistic noise fused model 203. The noise fused model 203 can then be used as training data 202 to train the weight analyzer 500/501 to learn to decode the noise 201.

[0239] FIG. 6 shows an exemplary representation of load in different aspects of railway infrastructure. Load data may be generated using weight sensor in the railway structure. The weight analyzer 500/501 may be configured to automatically learn at least one feature from the load data. Furthermore, model generator 300 may be configured to automatically generate a degradation model based on the load data. The processing component 100 may further be configured to automatically calculate a load factor, for example between 0.0 and 12.0, based on load data. In some embodiments wherein the load data is directly fed into the model analyzer 300, the model analyzer 300 may be configured to generate the load factor. The load factor may be based on number and/or type and/or speed of trains that passes over the sensor node. The database which may comprise a higher load factor may represent faster degradation. This degradation model can further assist in generating an inspection schedule of the railway infrastructure.

[0240] The weight analyzer 500/501 may further be configured to self-learn the at least one feature from the load data, such as quantity of gravel in the railway tracks.

[0241] Reference numbers and letters appearing between parentheses in the claims, identifying features described in the embodiments and illustrated in the accompanying drawings, are provided as an aid to the reader as an exemplification of the matter claimed. The inclusion of such reference numbers and letters is not to be interpreted as placing any limitations on the scope of the claims.

[0242] The term “at least one of a first option and a second option” is intended to mean the first option or the second option or the first option and the second option.

[0243] Whenever a relative term, such as “about”, “substantially” or “approximately” is used in this specification, such a term should also be construed to also include the exact term. That is, e.g., “substantially straight” should be construed to also include “(exactly) straight”.

[0244] Whenever steps were recited in the above or also in the appended claims, it should be noted that the order in which the steps are recited in this text may be accidental. That is, unless otherwise specified or unless clear to the skilled person, the order in which steps are recited may be accidental. That is, when the present document states, e.g., that a method comprises steps (A) and (B), this does not necessarily mean that step (A) precedes step (B), but it is also possible that step (A) is performed (at least partly) simultaneously with step (B) or that step (B) precedes step (A). Furthermore, when a step (X) is said to precede another step (Z), this does not imply that there is no step between steps (X) and (Z). That is, step (X) preceding step (Z) encompasses the situation that step (X) is performed directly before step (Z), but also the situation that (X) is performed before one or more steps (Y1), . . . , followed by step (Z). Corresponding considerations apply when terms like “after” or “before” are used.