Method and system for determining the load status of a carrying platform using radar

Abstract

A system for determining the load status of a platform includes a load status determination (LSD) device installed on the platform and includes a radar to emit waves and provide raw data based on reflected waves, a load estimator to use the raw data to compute a new score indicating the probability of a container being loaded, and a status determinator to estimate the load status based on the new score and a plurality of previous scores. A method for determining the load status of a platform includes emitting waves from a radar installed on the platform and providing raw data based on reflected waves, computing a new score indicating the probability of a container being loaded using the raw data, and determining the load status based on the new score and a plurality of previous scores.

Description

METHOD AND SYSTEM FOR DETERMINING THE LOAD STATUS OF A CARRYING PLATFORM USING RADARFIELD

[0001] The invention relates in general to determining the load status of a platform such as a carrying chassis, and more specifically utilizing a radar for this purpose.BACKGROUND

[0002] Container drayage, the process of transporting containers from a port or terminal to an inland location for unloading or further transport, as well as returning empty containers to the port or terminal for reuse, is a critical component of the logistics and supply chain process. It often serves as the initial or final leg of a container’s journey, facilitating the seamless movement of shipping containers between locations throughout the entire transportation process. Drayage typically requires specialized trucks equipped to handle standard shipping containers.

[0003] During this process, determining whether a container is present on the truck’s carrying chassis is essential for various logistical purposes and for fee and tax calculations. Collecting this data manually is complex and requires significant human resources.SUMMARY

[0004] There is provided in accordance with an embodiment of the invention, a system for determining the load status of a platform. The system includes a load status determination (LSD) device installed on the platform, wherein the LSD further includes a radar to emit electromagnetic waves and provide raw data based on reflected waves and at least one computer processor, a load estimator to use the raw data to compute a new probability score that indicates the likelihood of a container being loaded on the platform, and a status determinator to estimate the load status based on the new probability score and a plurality of previous probability scores, wherein the load estimator and the status determinator are operated the computer processor.

[0005] Additionally, in accordance with an embodiment of the invention, the LSD further includes a beam manipulation component to concentrate and direct the electromagnetic waves.

[0006] Moreover, in accordance with an embodiment of the invention, the load estimator further includes a data manipulator to split the raw data into segments related to the distance from the radar and to manipulate the raw data according to the 4thpower.

[0007] Furthermore, in accordance with an embodiment of the invention, the load estimator further includes a feature creation component to compute feature-engineered data items based on the raw data and wherein the load estimator to use the feature-engineered data items to compute the new probability score.

[0008] Stil further, in accordance with an embodiment of the invention, the feature- engineered data items comprise at least one of: a) maximum value of the raw data, b) a bin number of the maximum value, c) minimum value of the raw data, d) number of “peaks” in the raw data e) ratio between the maximum value and the minimum value, and f) average value of the raw data.

[0009] Additionally, in accordance with an embodiment of the invention, the load estimator further includes a machine learning model trained to compute the new probability score based on the raw data.

[0010] Moreover, in accordance with an embodiment of the invention, the machine learning model is further trained to compute the new probability score based on the feature- engineered data items.

[0011] Furthermore, in accordance with an embodiment of the invention, the load estimator machine learning model is random forest.

[0012] Still further, in accordance with an embodiment of the invention, the status determinator further includes a machine learning model trained to determine the load status based on the new probability score and the plurality of previous probability scores.

[0013] Moreover, in accordance with an embodiment of the invention, the load estimator is running on the LSD device and the status determinator is running on an application communicating with the LSD over a communications network.

[0014] There is provided in accordance with an embodiment of the invention, a method for determining the load status of a platform. The method includes emitting electromagnetic waves from a radar and providing raw data based on reflected waves, wherein the radar ispart of a load status determination (LSD) device installed on the platform, computing a new probability score that indicates the likelihood of a container being loaded on the platform using the raw data, and determining the load status based on the new probability score and a plurality of previous probability scores, wherein the computing step and the determining step are executed by at least one computer processor.

[0015] Additionally, in accordance with an embodiment of the invention, the method further includes concentrating and directing the electromagnetic waves.

[0016] Furthermore, in accordance with an embodiment of the invention, the method further includes manipulating the raw data into segments related to the distance from the radar and according to the 4thpower.

[0017] Still further, in accordance with an embodiment of the invention, the method further includes creating a plurality of feature-engineered data items based on the raw data and using the feature-engineered data items in the computing step.

[0018] Additionally, in accordance with an embodiment of the invention, the feature- engineered data items includes at least one of: a) maximum value of the raw data, b) a bin number of the maximum value, c) minimum value of the raw data, d) number of “peaks” in the raw data e) ratio between the maximum value and the minimum value, and f) average value of the raw data.

[0019] Moreover, in accordance with an embodiment of the invention, the computing step is performed using a machine learning model trained to compute the new probability score based on the raw data.

[0020] Furthermore, in accordance with an embodiment of the invention, the machine learning model is further trained to compute the new probability score based on the feature- engineered data items.

[0021] Still further, in accordance with an embodiment of the invention, the machine learning model is random forest.

[0022] Moreover, in accordance with an embodiment of the invention, the step of determining the load status is performed using a machine learning model trained to determine the load status based on the new probability score and the plurality of previous probability scores.

[0023] Additionally, in accordance with an embodiment of the invention, the step of computing a new probability score is performed on the LSD and the step of determining theload status is performed on an application communicating with the LSD over a communications network.BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The invention will now be described in relation to certain examples and embodiments with reference to the following illustrative drawing figures so that it may be more fully understood. In the drawings:Fig. 1 is a schematic illustration of a platform over which a container is loaded, and a load status determination (LSD) device designed to determine the status of the platform according to an embodiment of the invention;Fig. 2 is a schematic illustration of an LSD, constructed and operative in accordance with an embodiment of the invention;Fig. 3 is a schematic illustration of a surface and an area covered by radar’s projected signal on that surface according to an embodiment of the invention;Fig. 4 is a schematic illustration of a probability calculation program (PCP) constructed and operative in accordance with an embodiment of the invention;Fig. 5 is a schematic illustration of an application constructed and operative in accordance with an embodiment of the invention, configured to determine the status of the platform, andFig. 6 is a schematic illustration of a two-part flow used to determine the load status of the platform, according to an embodiment of the invention.

[0025] It will be appreciated that for simplicity and clarity of illustration, elements shown in the figures have not necessarily been drawn to scale. For example, the dimensions of some of the elements may be exaggerated relative to other elements for clarity. Further, where considered appropriate, reference numerals may be repeated among the figures to indicate the same or analogous elements.DETAILED DESCRIPTION

[0026] In the following description, various aspects of the invention will be described. For purposes of explanation, specific configurations and details are set forth in order to provide a thorough understanding of the invention. However, it will also be apparent to one skilled in the art that the invention may be practiced without the specific details presented herein. Furthermore, well-known features may be omitted or simplified in order not to obscure the invention.

[0027] Embodiments of the invention provide systems and methods for determining whether a carrying platform is loaded or unloaded, i.e. whether a container is present on the truck’s carrying chassis.

[0028] Fig. 1, to which reference is now made, is a schematic illustration of a platform 1 over which a container 2 may be loaded, and a load status determination (LSD) device 100, designed to determine whether container 2 is loaded on platform 1 according to an embodiment of the invention. LSD 100 may be installed on platform 1 and may communicate with an application 5 over a communications network 6.

[0029] Platform 1 may be part of any framework, vehicle, equipment, or chassis, such as a cart or dolly (e.g., a framework designed to carry and probably transport heavy equipment such as shipping containers), and container 2 may be any object designed to store and probably transport goods across various modes of transport, such as cargo units or shipping containers, or any other type of cargo that may be loaded on platform 1 . It may be noted that LSD 100 may be installed on any infrastructure used to carry any type of load. For simplicity, the terms ‘chassis’ and ‘platform’ may be used interchangeably to refer to the carrying structure, while the term ‘container’ may be used to refer to the carried load.

[0030] Communications network 6 may be any conventional means by which communications may be carried out, such as by a computer network like the Internet.

[0031] Application 5 may refer to a system designed to perform specific tasks or functions with the capability to communicate over communications network 6. Application 5 may be implemented using any computer hardware, software, or a combination of both in accordance with conventional techniques. Application 5 may incorporate a set of algorithms, including machine learning algorithms.

[0032] Application 5 may be responsible, among other things to inform the load status of platform 1 to a user. The load status may be determined based on information provided by LSD 100 and / or computed by application 5 using algorithms designed to enhance the accuracy of identifying the load status while minimizing the likelihood of false identification

[0033] Fig. 2, to which reference is now made, is a schematic illustration of LSD 100 constructed and implemented in accordance with an embodiment of the invention. LSD 100 comprises a beam manipulation component 210, a radar 220, a transceiver 230, a computer processor 240 and a data storage 250.

[0034] Transceiver 230 may be any combination of transmitters and receivers capable of communicating with communications devices over a wireless or wireline communication channel. Transceiver 230 may receive from application 5 information needed for its operation. The information may include commands, configuration, updated machine learning models (e.g., models that are used to determine the load status of the platform) and the like. Transceiver 230 may send information to application 5 including information received from the radar, information computed by LSD 100, the status of LSD 100 (e.g., battery, malfunctioning components etc.), the status of the platform (e.g., location) and any other parameters and information computed, sensed, detected and / or set by computer processor 240 or any other component withing LSD 100.

[0035] Computer processor 240 may be any known device capable of performing operations, calculations, and executing instructions. Computer processor 240 may use transceiver 230 to communicate with application 5 over communications network 6.

[0036] Data storage 250 may be any known device capable of storing digital information, such as data files, configurations and the like for later retrieval and may be designed to provide reliable, efficient, and secure access to stored data. Data storage 250 may include any known storage media such as hard drives, solid-state drives (SSDs), memory chips and the like.

[0037] Radar 220 may be a radar (Radio Detection and Ranging) system that uses electromagnetic waves (e.g., radio, microwave) to detect objects and measure their distance. Radar 220 may send out waves (typically radio), which bounce off objects in the environment, and provide estimates of reflected power per range bin (also referred to as range gate).

[0038] Radar 220 may be a low power radarthat consumes minimal electrical power. Using a low power design may ensure that radar 220 can operate for extended periods without needing frequent recharging or high energy input. To further reduce power consumption, radar 220 may be activated for short configurable periods (e.g., tens of seconds or minutes) in response to a physical or logical event, such as reaching a specific location, after a defined period, at specified intervals, upon receiving a request (e.g., from application 5), or similar triggers.

[0039] Radar 220 may receive the reflected power and may provide raw data that includes several data items as output. The data items may include a few (e.g., 8) numbers representing various characteristics of the reflection. It may be noted that the information provided by radar 220 may be noisy and may be influenced by weather conditions, the material of the loaded object, the angle of the object on platform 1 and the like.

[0040] Beam manipulation component 210 may comprise one or more lenses or a phase array of radiating elements used to concentrate and direct the electromagnetic waves created by radar 220 toward a specific location. For example, beam manipulation component 210 may be configured to direct the electromagnetic waves along the length of container 2, possibly creating an elliptic shape to cover a larger part of the bottom of container 2.

[0041] This configuration may facilitate the detection of various structural elements (e.g., members at the bottom of container 2), enabling the identification of container 2 on platform 1, as opposed to the presence of other items or the existence of noise.

[0042] All the various elements and components of LSD 100 may be encased in a polypropylene or similar material to enable proper functionality and operation.

[0043] Fig. 3, to which reference is now made, is a schematic illustration of a surface 310 (i.e., the bottom of container 2) and an area 330 covered by the radar’s projected signal on that surface. Surface 310 may be the floor panel of container 2 and may comprise multiple structural support members 320 distributed throughout its structure. Beam manipulation component 210 may manipulate beam 330 to an elliptic shape covering mainly bottom 310 with its support members 320. Beam manipulation component 210 may be configured to cover various parts of surface 310, enabling the detection of specific elements (structural support members 320) that identify a particular object (i.e., container 2).

[0044] Fig. 4, to which reference is now made, is a schematic illustration of a probability calculation program (PCP) 400 constructed and operative in accordance with an embodiment of the invention. PCP 400 may be operated by processor 240 and may comprise a data manipulator 405, a feature creation component 410, a probability assessment module 420 and a load estimation component 430.

[0045] Data manipulator 405 may receive the raw data provided by radar 220 and change it to improve the computation.

[0046] Data manipulator 405 may divide the raw data into segments based on the distance from radar 220, enabling correlation between different range segments and real load detection (e.g., radar 220 expects to see support members 320 in the reflected data from different range segments) rather than measured noise that tends not to correlate.

[0047] Data manipulator 405 may manipulate the raw data received from radar 220 according to the 4thpower of the distance between radar 220 and the reflection, as the signal power attenuates with the square of the distance traveled, and the reflection power is inversely proportional to the fourth power of the total distance (distance to the target and back).

[0048] Feature creation component 410 may receive the raw data from radar 220 and generate additional feature-engineered data items derived from it to enhance the accuracy of the load status prediction. The computed feature-engineered data items may include the following data items: the maximum value of the raw data provided by radar 220, the location (i.e., bin number) of the maximum value, the minimum value of the 8 outputs, the number of “peaks” in the raw data, the ratio between the maximum and minimum values, and the average value of the raw data.

[0049] Probability assessment module 420 may be a function designed to provide a probability score indicating the probability of platform 1 being loaded based on the raw data provided by radar 220, the probability score may be in the range from 0 to 1.

[0050] In one embodiment probability assessment module 420 may be a machine learning model trained to provide a probability score indicating the probability of platform 1 being loaded, based on the raw data and the system-engineered data items. The dataset for trainingprobability assessment module 420 may include as input labeled sets of raw data and feature- engineered data items (each set with an indication whether the platform is loaded or not).

[0051] The machine learning algorithm may be any suitable algorithm, including random forest, Bagging, Extra Trees and the like.

[0052] Load estimation component 430 may receive raw data from radar 220, send it to feature creation component 410 and receive the additional feature-engineered data items. Load estimation component 430 may then use the radar raw data and the feature-engineered data as input to probability assessment module 420 and receive a probability score indicating the probability of platform 1 being loaded with container 2.

[0053] Processor 240 may be configured to receive raw data from radar 220, activate load estimation component 430 to compute the probability score and transmit the probability score for further processing. The additional processing may be done by application 5 that may compute the current load status of platform 1 according to the information provided by LSD 100 over time. It may be noted that in an alternative embodiment, the additional processing may be done on LSD 100 and not on application 5.

[0054] Pig. 5, to which reference is now made, is a schematic illustration of application 5, constructed and operative in accordance with an embodiment of the invention, configured to determine the status of platform 1. Application 5 may comprise a data storage 510; a load status module 520 and a status determinator 530.

[0055] Data storage 510, (like data storage 250) may be any storage system designed to provide reliable, efficient, and secure access to stored data. Data storage 510 can be collocated with application 5 or located at another location over communications network 6. Data storage 510 may store probability scores received at different points in time and may retain these values as a time series.

[0056] Load status module 520 may be a function designed to determine the state of platform 1 based on current and past estimates. In one embodiment, load status module 520 may be a machine learning model trained to estimate the load status based on N probability scores representing the probability of container 2 being loaded on platform 1 in N consecutive reads of radar 220.

[0057] Status determinator 530 may use the last N probability scores as input to load status module 520 and obtain a decision regarding the load status of platform 1. The status of platform 1 may be loaded, unloaded and unknown. Using a machine learning model that uses time series to determine the load status of platform 1 relies on the assumption that the longer it takes to determine a status change, the higher the probability that the change really happened. However, a status change should be reported as soon as possible, therefore, the model may be designed to account for time to minimize false detections.

[0058] By considering current and past loads states of platform 1, load status module 520 may reflect real-world scenarios where the load status of platform 1 cannot flip instantly (e.g., transition from unloaded to loaded and then back to unloaded). This approach enables the model to avoid rapid, potentially erroneous status changes fluctuation, (e.g., keep the unloaded status).

[0059] Status determinator 530 may store the most recent probability score computed by LSD 100 in data storage 510 for future use.

[0060] Fig. 6, to which reference is now made, is a schematic illustration of a two-part flow 600 for determining the load status of platform 1, the flow constructed and operative in accordance with an embodiment of the invention.

[0061] The first part is a probability computation 610 that may compute a new probability score based on the current raw data provided by radar 220 and any additional feature- engineered data items. The second part is a status computation 650 that may use the latest N probability scores (including the latest one) to determine the current load status of platform 1.

[0062] The following steps may be included in probability computation 610. In step 615 the current radar raw data may be collected from radar 220. In step 620 the additional feature- engineered data items may be created using the radar raw data. In step 625 the raw data and the feature-engineered data may be used as input to the machine learning probability assessment module 420 that may compute a probability score representing the likelihood of a load (e.g., container 2) being placed on platform 1 and in step 630 the probability score may be maintained for further processing.

[0063] The following steps may be included in status computation 650. In step 655 the latest N probability scores may be used as input to load status module 520 to estimate the currentload status and in step 660, if the load status has changed, transmit the current load status to a user.

[0064] It is worth noting that entire flow 600 (both probability computation 610 and status computation 650) may be executed on LSD 100 or on application 5 or may be split between the two (i.e., probability computation 610 executed on LSD 100 and status computation 650 executed on application 5). The specific implementation depends on the requirements of each installation.

[0065] When executed on LSD 100 all the information required to perform the computation (including both machine learning models) may be stored in data storage 250 and executed by processor 240. When executed on application 5, all the information may be stored in data storage 510 and executed by the machine on which it is installed or deployed, and the raw data collected by radar 220 may be transmitted to application 5 over communication networks 6.

[0066] While certain features of the invention have been illustrated and described herein, many variations, modifications, substitutions, changes, and equivalents will now occur to those of ordinary skill in the art.

[0067] It may be appreciated that the steps shown for the flows herein above are not intended to be limiting and that each flow may be practiced with variations. These variations may include more steps, less steps, changing the sequence of steps, skipping steps, among other variations which may be evident to one skilled in the art.

[0068] Unless specifically stated otherwise, as apparent from the following discussions, it is appreciated that throughout the specification discussions utilizing terms such as “analyzing”, “processing,” “computing,” “calculating,” “determining,” “detecting”, “identifying” or the like, refer to the action and / or processes of a computer or computing system, or similar electronic computing device, that manipulates and / or transforms data represented as physical, such as electronic, quantities within the computing system’s registers and / or memories into other data similarly represented as physical quantities within the computing system’s memories, registers or other such information storage, transmission or display devices.

Claims

CLAIMSWhat is claimed is:

1. A system for determining a load status of a platform, the system comprising: a load status determination (LSD) device installed on the platform, wherein the LSD further comprises a radar to emit electromagnetic waves and provide raw data based on reflected waves and at least one computer processor; a load estimator to use the raw data to compute a new probability score that indicates the likelihood of a container being loaded on the platform; and a status determinator to estimate the load status based on the new probability score and a plurality of previous probability scores, wherein the load estimator and the status determinator are operated by the at least one computer processor.

2. The system of claim 1 wherein the LSD further comprises a beam manipulation component to concentrate and direct the electromagnetic waves.

3. The system of claim 1 wherein the load estimator further comprises a data manipulator to split the raw data into segments related to the distance from the radar and to manipulate the raw data according to the 4thpower.

4. The system of claim 1 wherein the load estimator further comprises a feature creation component to compute feature-engineered data items based on the raw data and wherein the load estimator to use the feature-engineered data items to compute the new probability score.

5. The system of claim 4 wherein the feature-engineered data items comprise at least one of: a) maximum value of the raw data, b) a bin number of the maximum value, c) minimum value of the raw data, d) number of “peaks” in the raw data e) ratio between the maximum value and the minimum value, and f) average value of the raw data.

6. The system of claim 1 wherein the load estimator further comprises a machine learning model trained to compute the new probability score based on the raw data.

7. The system of claim 6 wherein the machine learning model is further trained to compute the new probability score based on the feature-engineered data items.

8. The system of claim 6 wherein the load estimator machine learning model is random forest.

9. The system of claim 1 wherein the status determinator further comprises a machine learning model trained to determine the load status based on the new probability score and the plurality of previous probability scores.

10. The system of claim 1 wherein the load estimator is running on the LSD device and the status determinator is running on an application communicating with the LSD over a communications network.

11. A method for determining a load status of a platform, the method comprising: emitting electromagnetic waves from a radar and providing raw data based on reflected waves, wherein the radar is installed on the platform; computing a new probability score that indicates the likelihood of a container being loaded on the platform using the raw data; and determining the load status based on the new probability score and a plurality of previous probability scores, wherein the computing step and the determining step are executed by at least one computer processor.

12. The method of claim 11 further comprising concentrating and directing the electromagnetic waves.

13. The method of claim 11 further comprising manipulating the raw data into segments related to the distance from the radar and according to the 4thpower.

14. The method of claim 11 further comprising creating a plurality of feature-engineered data items based on the raw data and using the feature-engineered data items in the computing step.

15. The method of claim 14 wherein the feature-engineered data items comprise at least one of: a) maximum value of the raw data, b) a bin number of the maximum value, c) minimum value of the raw data, d) number of “peaks” in the raw data e) ratio between the maximum value and the minimum value, and f) average value of the raw data.

16. The method of claim 11 wherein the computing step is performed using a machine learning model trained to compute the new probability score based on the raw data.

17. The method of claim 16 wherein the machine learning model is further trained to compute the new probability score based on the feature-engineered data items.

18. The method of claim 16 wherein the machine learning model is random forest.

19. The method of claim 11 wherein the step of determining the load status is performed using a machine learning model trained to determine the load status based on the new probability score and the plurality of previous probability scores.

20. The method of claim 11 wherein the step of computing a new probability score is performed on a device installed on the platform and the determining the load status step is performed on an application communicating with the device over a communications network.

Images