Methods and systems for predicting an energy consumption of a vessel

Abstract

Methods and server systems for predicting energy consumption of a vessel are described herein. The method performed by a server system includes accessing a set of stable vessel operating parameters from a plurality of vessel operating parameters of a vessel, recorded at predefined intervals for the vessel. Herein, the set of stable vessel operating parameters satisfies stability criteria. The method further includes determining a subset of stable vessel operating parameters from the set of stable vessel operating parameters. Herein, each stable vessel operating parameter in the subset of stable vessel operating parameters satisfies a performance threshold. The method further includes generating a set of features based on the subset of stable vessel operating parameters. The method further includes predicting, by a Machine Learning (ML) model, an energy consumption for the vessel based on applying the set of features on the ML model.

Description

METHODS AND SYSTEMS FOR PREDICTING AN ENERGY CONSUMPTION OFA VESSELTECHNICAL FIELD

[0001] The present disclosure relates to the field of predicting an energy consumption of a vessel and, more particularly, to electronic methods and complex processing systems for predicting the energy consumption of an engine such as the main or auxiliary engine of the vessel.BACKGROUND

[0002] The shipping industry is known as the backbone of global trade or commerce. This global trade is quite volatile and relies on the shipping industry to maintain prices for various commodities across the globe. Therefore, the shipping industry must operate efficiently through cost-effective operations to ensure stability in global commodity prices. An important aspect of measuring this efficiency is energy consumption. For instance, energy consumption in the form of fuel consumption is one of the key metrics used by the shipping industry in regulating their maritime operations. Generally, in the shipping industry, energy consumption is computed in terms of Specific Fuel Oil Consumption (SFOC). The term ‘Specific Fuel Oil Consumption’ or ‘SFOC’ can be defined as the amount of fuel consumed per unit of power output from the engine such as the main or auxiliary engine of a vessel (or a maritime vessel). SFOC is measured in grams per kilowatt-hour (g / kWh). As may be understood, to maintain efficiency during operations, operators of vessels have to compute or predict the energy consumption of their vessels for all planned routes. This computation or prediction is performed to select efficient routes out of different possible routes for a shipping journey. Thus, accurate prediction or management of the energy consumption or SFOC is essential for optimizing fuel usage, improving engine performance, and minimizing the operational costs of the vessels.

[0003] Conventionally, the prediction of energy consumption for a vessel is performed manually. This manual process often relies on historical data and empirical assumptions. In some instances, the operators collect fuel consumption data and engine performance data over time and use this data to estimate the SFOC values. These calculations generally rely on various assumptions regarding the engine’s operating conditions, load, environmental factors, and so on. However, these conventional approaches for determining the energy consumptionP24-009PCT1of the vessel present various challenges. It is noted that since the conventional predictions are performed using historical data, they do not reflect the real-time variations in engine performance, fuel mix, fuel quality, environmental conditions, and so on. Further, these predictions are susceptible to human error, and inaccuracies due to improper data interpretation as well.

[0004] These challenges or deficiencies result in poor energy consumption predictions which impact the selection of the proposed route for a maritime journey. This in turn may lead to fuel wastage causing financial loss and environmental degradation along with possible hazardous operating conditions for the vessel due to improper route selection.

[0005] Thus, it is desirable to find technological solutions predicting the energy consumption of an engine such as the main or auxiliary engine of the vessel.SUMMARY

[0006] There exists a need for techniques to overcome one or more limitations stated above such as the adverse impact of poor energy consumption predictions on the operating conditions of a vessel, fuel wastage, improper route selection, and so on.

[0007] Various embodiments of the present disclosure provide methods and systems for predicting the energy consumption of an engine such as the main or auxiliary engine of the vessel automatically while relying on near or real-time data (z.e., vessel operating parameters) for the said prediction. Various embodiments of the present disclosure describe a computing device and a method that helps to determine the fuel wastage by the vessel.

[0008] To achieve the above and other objectives of the present disclosure, in one aspect, a computer-implemented method for predicting the energy consumption of a vessel is disclosed. The computer-implemented method is performed by a server system. The computer- implemented method includes accessing a set of stable vessel operating parameters from a plurality of vessel operating parameters of a vessel, recorded at predefined intervals for the vessel. Herein, the set of stable vessel operating parameters satisfies stability criteria. The computer-implemented method further includes determining a subset of stable vessel operating parameters from the set of stable vessel operating parameters. Herein, each stable vessel operating parameter in the subset of stable vessel operating parameters satisfies a performance threshold. The computer-implemented method further includes generating a set of features based, at least in part, on the subset of stable vessel operating parameters. The computer- implemented method further includes predicting, by a Machine Learning (ML) model, anP24-009PCT1energy consumption for the vessel based, at least in part, on applying the set of features on the ML model.

[0009] An advantage of some embodiments is that by accessing the set of stable vessel parameters from the plurality of vessel operating parameters, the prediction process for the energy consumption can be performed for the stable operating conditions vessel. This aspect ensures that temporary transients, weather conditions, etc., among other temporary unstable conditions, don’t affect the predictions. Another advantage of some embodiments is that by extracting the subset of stable vessel operating parameters from the set of stable vessel operating parameters using the performance threshold in the process of predicting the energy consumption, the server system ensures that the predicted energy consumption represents energy consumption during the top or best operating state of the vessel.

[0010] In an aspect, the computer-implemented method further includes accessing a historical vessel performance dataset including a historical set of stable vessel operating parameters associated with each vessel of a plurality of vessels from a database. The computer- implemented method further includes determining a historical subset of stable vessel operating parameters from the historical set of stable vessel operating parameters for the each vessel. Herein, each historical stable vessel operating parameter in the historical subset of stable vessel operating parameters satisfies the performance threshold. The computer-implemented method further includes training, the ML model based, at least in part, on the historical subset of stable vessel operating parameters.

[0011] An advantage of some embodiments is that by training the ML model using the historical subset of stable vessel operating parameters, the ML model learns from the historical data to understand the underlying hidden patterns or dependencies to perform accurate predictions.

[0012] In an aspect, the step of training the ML model includes splitting the historical subset of stable vessel operating parameters for the each vessel into a training set and a validation set. The step further includes iteratively performing following set of operations till training criteria are met: (1) initializing the ML model based, at least in part, on a plurality of hyperparameters, wherein the ML model is an ensemble ML model configured to perform quantile regression with a quantile parameter set to the performance threshold, (2) generating a set of historical features based, at least in part, on the training set, (3) generating, by the ML model, an energy consumption prediction for at least one time interval associated with theP24-009PCT1historical subset of stable vessel operating parameters in the validation set, based, at least in part, on applying the set of historical features on the ML model, (4) computing, by the ML model using a quantile loss function, at least one quantile loss between the energy consumption prediction and an actual energy consumption computed using the validation set for the at least one time interval, and (5) optimizing the quantile loss function based, at least in part, on the at least one quantile loss to update the plurality of hyperparameters.

[0013] An advantage of some embodiments is that configuring the ensemble ML model to perform quantile regression with a quantile parameter set to the performance threshold allows the ML model to predict the energy consumption for the top or best quantile. Further, relying on the quantile loss function for optimizing the ML model improves the performance of the said model as well. Further, since the ML model is trained using data associated with the plurality of vessels, the trained ML model will be applicable for performing predictions related to any vessel of the plurality of vessels.

[0014] In an aspect, the step of accessing the set of stable vessel operating parameters includes recording the plurality of vessel operating parameters from at least one data source associated with the vessel at one or more frequencies. The step further includes aggregating the plurality of recorded vessel operating parameters at the predefined intervals. The step further includes extracting the set of stable vessel operating parameters from the plurality of aggregated vessel operating parameters based, at least in part, on the stability criteria associated with the vessel. Herein, the stability criteria define stable operating conditions for the vessel.

[0015] An advantage of some embodiments is that aggregating the plurality of vessel operating parameters recorded at one or more frequencies into predefined intervals improves the data processing efficiency of the ML model. Further, extracting stable vessel operating parameters ensures that the prediction process for the energy consumption can be performed for the stable operating conditions vessel. As described earlier, this aspect ensures that temporary transients, weather conditions, etc., among other temporary unstable conditions, don’t affect the predictions.

[0016] In an aspect, the step of extracting the set of stable vessel operating parameters includes identifying one or more invalid vessel operating parameters from the plurality of vessel operating parameters based, at least in part, on first filtering criteria within the stability criteria. Herein, the first filtering criteria is based, at least in part, on at least one of a stable engine power range, a stable shaft Revolutions Per Minute (RPM) range, a stable engine loadP24-009PCT1range, a stable fuel consumption range, or a Speed Over Ground (SOG) range. The step further includes identifying one or more frozen vessel operating parameters from the plurality of vessel operating parameters based, at least in part, on second filtering criteria within the stability criteria. Herein, the second filtering criteria is based, at least in part, on at least one of a stagnant engine power factor, a stagnant shaft RPM factor, or a stagnant fuel consumption factor. The step further includes eliminating the one or more invalid vessel operating parameters and the one or more frozen vessel operating parameters from the plurality of vessel operating parameters to determine a set of remaining vessel operating parameters.

[0017] An advantage of some embodiments is that the plurality of vessel operating parameters can be filtered to remove the invalid vessel operating parameters and the frozen vessel operating parameters. This aspect improves the efficiency of the ML model during the prediction process by eliminating noise from the set of remaining vessel operating parameters.

[0018] In an aspect, the step of extracting the set of stable vessel operating parameters further includes filtering the set of stable vessel operating parameters from the set of remaining vessel operating parameters based, at least in part, on third filtering criteria within the stability criteria. Herein, the third filtering criteria is based, at least in part, on at least one of a stable engine load range, at least one stable shaft RPM operating range, a stable true heading range, a stable rudder angle range, or a stable SOG operating range.

[0019] An advantage of some embodiments is that the set of remaining vessel operating parameters can be filtered to obtain the set of stable vessel operating parameters. This aspect improves the efficiency of the ML model during the prediction process by eliminating noise from the set of stable vessel operating parameters.

[0020] In an aspect, the step of determining the subset of stable vessel operating parameters includes accessing the performance threshold indicating a selection quantile. The step further includes identifying and extracting the subset of vessel operating parameters present within the selection quantile from the set of stable vessel operating parameters.

[0021] An advantage of some embodiments is that using the performance threshold allows the ML model to predict the energy consumption for the top or best quantile.

[0022] In an aspect, the computer-implemented method further includes recording a plurality of actual vessel operating parameters from at least one data source. The computer- implemented method further includes computing an actual energy consumption based, at least in part, on the plurality of actual vessel operating parameters. The computer-implementedP24-009PCT1method further includes comparing the actual energy consumption with the predicted energy consumption. The computer-implemented method further includes determining a fuel wastage by the vessel based, at least in part, on the comparison step.

[0023] An advantage of some embodiments is that by determining the fuel wastage by the vessel an operator of the said vessel can ascertain how he / she has deviated from the predicted energy consumption. This deviation provides an indication to the operator that there is a need of maintenance and thereby improve the performance of the vessel. In other words, this understanding may help the operator in improving the efficiency of the vessel during upcoming portions of shipping journey.

[0024] In an aspect, the computer-implemented method further includes facilitating the generation of at least one Graphical User Interface (GUI) based, at least in part, on the comparing step. Herein, the at least one GUI provides a visualization of the comparison between the actual energy consumption with the predicted energy consumption.

[0025] An advantage of some embodiments is that the GUI allows the operator to interact with the server system and understand the prediction and actual energy consumption using visualization.

[0026] As per another embodiment of the present disclosure, a server system is disclosed. The server system includes a communication interface and a memory including executable instructions. The server system also includes a processor communicably coupled to the memory. The processor is configured to execute the instructions to cause the server system, at least in part, to access a set of stable vessel operating parameters from a plurality of vessel operating parameters of a vessel, recorded at predefined intervals for the vessel. Herein, the set of stable vessel operating parameters satisfies stability criteria. The server system is further caused to determine a subset of stable vessel operating parameters from the set of stable vessel operating parameters. Herein, each stable vessel operating parameter in the subset of stable vessel operating parameters satisfies a performance threshold. The server system is further caused to generate a set of features based, at least in part, on the subset of stable vessel operating parameters. The server system is further caused to predict, by a Machine Learning (ML) model, an energy consumption for the vessel based, at least in part, on applying the set of features on the ML model.

[0027] As per yet another embodiment of the present disclosure, a non-transitory computer-readable storage medium is disclosed. The non-transitory computer-readable storageP24-009PCT1medium includes computer-executable instructions that, when executed by at least a processor of a server system, cause the server system to perform a method. The method includes accessing a set of stable vessel operating parameters from a plurality of vessel operating parameters of a vessel, recorded at predefined intervals for the vessel. Herein, the set of stable vessel operating parameters satisfies stability criteria. The method further includes determining a subset of stable vessel operating parameters from the set of stable vessel operating parameters. Herein, each stable vessel operating parameter in the subset of stable vessel operating parameters satisfies a performance threshold. The method further includes generating a set of features based, at least in part, on the subset of stable vessel operating parameters. The method further includes predicting, by a Machine Learning (ML) model, an energy consumption for the vessel based, at least in part, on applying the set of features on the ML model.

[0028] The foregoing summary is illustrative only and is not intended to be in any way limiting. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features will become apparent by reference to the drawings and the following detailed description.BRIEF DESCRIPTION OF FIGURES

[0029] For a more complete understanding of example embodiments of the present technology, reference is now made to the following descriptions taken in connection with the accompanying drawings in which:

[0030] FIG. 1 is an example representation of a maritime environment, in accordance with various embodiments of the present disclosure;

[0031] FIG. 2 illustrates a simplified block diagram of a server system, in accordance with an embodiment of the present disclosure;

[0032] FIG. 3 illustrates a schematic representation of the process for predicting an energy consumption of a vessel, in accordance with an embodiment of the present disclosure;

[0033] FIG. 4 illustrates a flow diagram of a method of operating the server system for training the Machine Learning (ML) model for predicting the energy consumption of a vessel, in accordance with an embodiment of the present disclosure; and

[0034] FIG. 5 illustrates a flow diagram of a method of operating the server system for predicting energy consumption for a vessel, in accordance with an embodiment of the presentP24-009PCT1disclosure.

[0035] The drawings referred to in this description are not to be understood as being drawn to scale except if specifically noted, and such drawings are only exemplary in nature.DETAILED DESCRIPTION

[0036] In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the present disclosure. It will be apparent, however, to one skilled in the art that the present disclosure can be practiced without these specific details. Descriptions of well-known components and processing techniques are omitted to not obscure the embodiments herein unnecessarily. The examples used herein are intended merely to facilitate an understanding of ways in which the embodiments herein may be practiced and to further enable those of skill in the art to practice the embodiments herein. Accordingly, the examples should not be construed as limiting the scope of the embodiments herein.

[0037] References in this specification to “one embodiment” or “an embodiment” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present disclosure. The appearances of the phrase “in an embodiment” in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Moreover, various features are described which may be exhibited by some embodiments and not by others. Similarly, various requirements are described which may be requirements for some embodiments but not for other embodiments.

[0038] Moreover, although the following description contains many specifics for the purposes of illustration, anyone skilled in the art will appreciate that many variations and / or alterations to said details are within the scope of the present disclosure. Similarly, although many of the features of the present disclosure are described in terms of each other, or in conjunction with each other, one skilled in the art will appreciate that many of these features can be provided independently of other features. Accordingly, this description of the present disclosure is set forth without any loss of generality to, and without imposing limitations upon, the present disclosure.

[0039] Conditional language such as, among others, "can," "could," "might" or "may," unless specifically stated otherwise, are otherwise understood within the context as used inP24-009PCT1general to convey that certain embodiments include, while other embodiments do not include, certain features, elements and / or steps. Thus, such conditional language is not generally intended to imply that features, elements, and / or steps are in any way required for one or more embodiments or that one or more embodiments necessarily include logic for deciding, with or without user input or prompting, whether these features, elements and / or steps are included or are to be performed in any particular embodiment.

[0040] Disjunctive language such as the phrase "at least one of X, Y, or Z" unless specifically stated otherwise, is otherwise understood with the context as used in general to present that an item, term, etc., may be either X, Y, or Z, or any combination thereof (e.g., X, Y, and / or Z). Thus, such disjunctive language is not generally intended to, and should not, imply that certain embodiments require at least one of X, at least one of Y, or at least one of Z to each be present.

[0041] Unless otherwise explicitly stated, articles such as "a" or "an" should generally be interpreted to include one or more described items. Accordingly, phrases such as "a server system configured to" are intended to include one or more recited server systems / processors. Such one or more recited devices can also be collectively configured to carry out the stated recitations. For example, "a processor configured to carry out recitations A, B, and C" can include a first processor configured to carry out recitation A working in conjunction with a second processor configured to carry out recitations B and C. The same holds true for the use of definite articles used to introduce embodiment recitations. In addition, even if a specific number of an introduced embodiment recitation is explicitly recited, those skilled in the art will recognize that such recitation should typically be interpreted to mean at least the recited number (e.g., the bare recitation of "two recitations," without other modifiers, typically means at least two recitations or two or more recitations).

[0042] It will be understood by those within the art that, in general, terms used herein, are generally intended as "open" terms (e.g., the term "including" or “comprising” should be interpreted as "including / comprising but not limited to," the term "having" should be interpreted as "having at least," the term "includes" or “comprises” should be interpreted as "includes / comprises but is not limited to," etc.).

[0043] For expository purposes, the term ‘vessel’, ‘boat’, ‘ship’, or ‘carrier’ (used interchangeably herein) refers to any type of vehicle or craft that is designed to navigate or operate in a fluid such as but not limited to water. Examples of vessels include commercialP24-009PCT1vessels, recreational vessels, special purpose vessels, and so on.

[0044] FIG. 1 is an example representation of a maritime environment 100, in accordance with various embodiments of the present disclosure. The maritime environment 100 includes a server system 102, a vessel 104, and one or more data sources 106, each coupled to, and in communication with (and / or with access to) a network 108. The vessel 104 may be, but is not limited to a maritime vessel, an aircraft, a boat, a ship, a yacht, a commercial cargo ship, and so on.

[0045] The network 108 may include, without limitation, a Light Fidelity (Li-Fi) network, a Local Area Network (LAN), a Wide Area Network (WAN), a Metropolitan Area Network (MAN), a satellite network, the Internet, a fiber optic network, a coaxial cable network, an Infrared (IR) network, a Radio Frequency (RF) network, a virtual network, and / or another suitable public and / or private network capable of supporting communication among two or more of the parts or components illustrated in FIG. 1, or any combination thereof.

[0046] Various entities in the maritime environment 100 may connect to the network 108 in accordance with various wired and wireless communication protocols, such as Transmission Control Protocol / Internet Protocol (TCP / IP), User Datagram Protocol (UDP), 2nd Generation (2G), 3rd Generation (3G), 4th Generation (4G), 5th Generation (5G) communication protocols, Long Term Evolution (LTE) communication protocols, future communication protocols or any combination thereof. For example, the network 108 may include multiple different networks, such as a private network made accessible by the server system 102 and a public network (e.g., the Internet, etc.) through which the server system 102, the vessel 104, and the one or more data sources 106 may communicate.

[0047] The vessel 104 may be operating in a sea following a predetermined route. The vessel 104 may be enabled with the Internet of Things (loT). In other words, the vessel 104 may be associated with one or more data sources 106 that collect, transmit, and analyze data in real-time. The integration of loT technology in the vessel 104 allows for seamless communication between various components and provides actionable insights to the operator of the vessel 104.

[0048] Examples of one or more data sources 106 may include but are not limited to engine and machinery sensors, navigation systems, environmental sensors / sy stems, hull monitoring systems, fuel management systems, communication systems, and so on. In an example, the engine and machinery sensors may be responsible for recording / monitoringP24-009PCT1parameters such as engine temperature, fuel consumption, RPM (revolutions per minute), oil pressure, coolant levels, and so on. The Navigation systems may be responsible for recording / monitoring parameters such as Global Positioning System (GPS), Radio Detection And Ranging (RADAR), Sound Navigation and Ranging (SONAR), and Automatic Identification Systems (AIS) to track the vessel’s location, speed, heading, rudder angle, surrounding marine traffic, and so on. The environmental sensors / sy stems may be responsible for recording / monitoring parameters such as external environmental conditions, including sea state (wave height and frequency), wind speed and direction, air and water temperature, humidity, and barometric pressure, among other weather conditions. In some instances, the environmental systems may access weather-related information from different meteorological departments or the internet as well. The hull monitoring systems may include strain gauges and accelerometers placed on the hull of the vessel 104 to record / monitor parameters such as stress, strain, and vibrations of the hull. The fuel management systems may be responsible for recording / monitoring parameters such as fuel levels, consumption rates, fuel mix type, fuel quality, and so on. The Satellite and radio communication systems are responsible for collecting and transmitting data between the vessel 104 and onshore operations centers.

[0049] In an embodiment, the one or more data sources 106 may be responsible for collecting or recording a plurality of vessel operating parameters of the vessel 104. The one or more data sources 106 may include a combination of sensors, onboard systems, and external data feeds, all integrated to provide comprehensive monitoring and data collection of the various vessel operating parameters.

[0050] In an instance, the one or more data sources 106 record the vessel operating parameters at one or more frequencies. For instance, the set of vessel operating parameters may be recorded every few milliseconds, seconds, minutes, or so on. In another instance, the data recording process for a few vessel operating parameters may take place using high- frequency medium (every few milliseconds to seconds), medium -frequency recording (every few minutes to hours), or low-frequency recording (every few hours or days) as well.

[0051] Examples of the vessel operating parameters include, but are not limited to, engine power, shaft Revolutions Per Minute (RPM), engine load, engine room temperature, pressure within engine room, power generated by the waste heat recovery system's steam and power turbines, inlet and outlet temperatures of the exhaust gas for turbo charger, temperature and pressure in the scavenging air receiver of main engine and / or auxiliary engine, pressure in the exhaust receiver of main engine and / or auxiliary engine, back pressure of exhaust gas inP24-009PCT1turbo charger, temperature and viscosity of the fuel oil, average temperature of the exhaust gas, maximum continuous rating, maximum shaft RPM, whether the main engine and / or auxiliary engine has a turbocharger cut-out, number of turbochargers, number of cylinders, diameter of each of the cylinders, number of piston strokes, length of the piston stroke, hull capacity for refrigerated cargo and dimensions, fuel consumption, Speed Over Ground (SOG), true heading, rudder angle, data on weather conditions, sea state, data on cargo conditions, ballast tank levels, overall fuel levels, fuel type, fuel mix (of mixed fuel batch), hull integrity, machinery condition, distance traveled, estimated time of arrival, etc., among other suitable vessel operating parameters. Since these vessel operating parameters are recorded at different frequencies (or the same high frequency), these parameters are aggregated at predefined intervals to enable simplified processing. Examples of the predefined intervals include 5 minutes (min.), 10 min., 15 min., and so on.

[0052] As described earlier, energy consumption prediction plays a key role in optimizing fuel usage, improving engine performance, and minimizing the operational costs of the vessels. Further, the energy consumption prediction also enables the operator to select an efficient route from different available route options as well. However, due to the conventional approach of performing these predictions manually using historical data, the prediction process is highly inefficient and ridden with inaccuracies as well.

[0053] To overcome this problem, an approach for predicting an energy consumption of the vessel 104 is required. To that end, to address the above-mentioned limitation, the present disclosure describes that the server system 102 may predict the energy consumption of the vessel 104.

[0054] In one embodiment, the maritime environment 100 may further include a database 110 coupled with the server system 102. In an example, the server system 102 coupled with the database 110 is embodied within a central server (not shown) associated with the operator of the vessel 104, however, in other examples, the server system 102 can be a standalone component (acting as a hub) connected to the central server. The database 110 may be incorporated in the server system 102 or maybe an individual entity connected to the server system 102 or maybe a database stored in cloud storage. In one embodiment, the database 110 may store the vessel operating parameters recorded by the one or more data sources 106, an Artificial Intelligence (Al) or Machine Learning (ML) model, and other necessary machine instructions required for implementing the various functionalities of the server system 102 such as firmware data, operating system, and the like. It is noted that the ML model has beenP24-009PCT1explained in detail later in the present disclosure. In addition, the database 110 provides a storage location for data and / or metadata obtained from various operations performed by the server system 102.

[0055] In an embodiment, the server system 102 is configured to access a set of stable vessel operating parameters from a plurality of vessel operating parameters of the vessel 104. Herein, the set of stable vessel operating parameters is selected from the plurality of vessel operating parameters such that each of these parameters satisfies stability criteria. It is noted that the stability criteria are predefined by an administrator (not shown) associated with the server system 102. In an instance, the stability criteria include a set of predefined operating conditions or a set of stable operating conditions for the vessel 104. In other words, the stability criteria define conditions during which the vessel operating parameters are considered stable thus, free of noisy, frozen, and / or invalid values.

[0056] In another embodiment, the server system 102 is configured to determine a subset of stable vessel operating parameters from the set of stable vessel operating parameters. Herein, each stable vessel operating parameter in the subset of stable vessel operating parameters is selected such that the said stable vessel operating parameter satisfies a performance threshold. In an implementation, the performance threshold is predefined by the administrator associated with the server system 102. For example, the performance threshold may be set to the top 20% to extract the top 20% of the vessel operating parameters within the set of stable vessel operating parameters to form the subset of stable vessel operating parameters. It is noted that the performance threshold can be configured as per the requirements of the operator of the vessel 104. Then, the server system 102 is configured to generate a set of features based, at least in part, on the subset of stable vessel operating parameters. The term ‘feature’ refers to variables or attributes that describe the data such as the vessel operating parameters. In particular, features serve as a way to represent the data in a format that can be understood and processed by Al or ML models. It is noted that one or more feature generation techniques (known as feature engineering) such as, but not limited to, time-based feature generation, binning, aggregated variable generation, dimensionality reduction, one hot encoding, frequency encoding, and so on, may be used for generating the features. Since these operations are well known in the art, the same have not been described herein for the sake of brevity.

[0057] In another embodiment, the server system 102 is configured to predict an energy consumption for the vessel 104. In particular, the server system 102 is configured toP24-009PCT1predict the energy consumption based, at least in part, on applying the set of features on the ML model. In an implementation, the ML model may be configured to predict Specific Fuel Oil Consumption (SFOC) value for the main or auxiliary engine of the vessel 104. In such an implementation, the SFOC value may correspond to the energy consumption of the vessel 104. It is noted that even though the SFOC value is used to describe the energy consumption of the vessel 104, the same should not be construed as a limitation of the present disclosure. For instance, instead of an SFOC value, the server system 102 may be configured to predict Engine Fuel Efficiency (EFE), Engine Power Output (EPO), Total Fuel Consumption (TFC), Energy Consumption per Nautical Mile (kWh / NM), and so on may be predicted to ascertain the energy consumption of the vessel 104 as well. A detailed explanation of various operations required for predicting the energy consumption of the vessel 104 by the server system 102 is provided later in reference to FIG. 2.

[0058] According to an aspect, the energy consumption is predicted or computed for a specific load associated with the vessel 104. In some instances, the energy consumption can be normalized using a standard energy density. This aspect ensures that the computed SFOC can be applied to different fuel types as well. As may be understood, each fuel type is associated with its corresponding energy density, thus the normalized SFOC can adjusted according to different energy densities for various fuel types for determining the SFOC of the vessel 104 operating on the various fuel types. Further, the SFOC can be computed for vessels operating using fuel mixes as well. In particular, Lower calorific value (LCV) correction is performed to adjust for different types of fuel being utilized by the vessel 104. This aspect allows the ML model to correct for the quality of fuel in terms of efficiency.

[0059] Although in Fig. 1, the server system 102 is shown to be incorporated within the maritime environment 100, in some embodiments, the server system 102 may be external to and in communication with the maritime environment 100, for example, via the network 108. In some examples, the server system 102 may be implemented in third-party external servers to perform the various operations described herein.

[0060] The number and arrangement of systems, devices, and / or networks shown in FIG. 1 are provided as an example. There may be additional systems, devices, and / or networks; fewer systems, devices, and / or networks; different systems, devices, and / or networks; and / or differently arranged systems, devices, and / or networks than those shown in FIG. 1. Furthermore, two or more systems or devices shown in FIG. 1 may be implemented within a single system or device, or a single system or device is shown in FIG. 1 may be implementedP24-009PCT1as multiple, distributed systems or devices. In addition, the server system 102 should be understood to be embodied in at least one computing device in communication with the network 108, which may be specifically configured, via executable instructions, to perform steps as described herein, and / or embodied in at least one non-transitory computer-readable media.

[0061] FIG. 2 illustrates a simplified block diagram of a server system 200, in accordance with an embodiment of the present disclosure. It is noted that the server system 200 may be similar to the server system 102 of FIG. 1. In one embodiment, the server system 200 is a part of the internal server operated by an organization employing the operator or an onboard personnel of the vessel 104. In some embodiments, the server system 200 is embodied as a cloud-based and / or Software as a Service (SaaS) based architecture.

[0062] The server system 200 includes a computer system 202 and a database 204. It is noted that the database 204 is identical to the database 110 of FIG. 1. The computer system 202 includes at least one processor 206 (herein, referred to interchangeably as ‘processor 206’) for executing instructions, a memory 208, a communication interface 210, a user interface 212 and a storage interface 214 that communicates with each other via a bus 216.

[0063] In some embodiments, the database 204 is integrated into the computer system 202. For example, the computer system 202 may include one or more hard disk drives as the database 204. A storage interface 214 is any component capable of providing the processor 206 with access to the database 204. The storage interface 214 may include, for example, an Advanced Technology Attachment (ATA) adapter, a Serial ATA (SATA) adapter, a Small Computer System Interface (SCSI) adapter, a RAID controller, a SAN adapter, a network adapter, and / or any component providing the processor 206 with access to the database 204. In one non-limiting example, the database 204 is configured to store a historical vessel performance dataset 218, a Machine Learning (ML) model 220, and the like.

[0064] In an example, the historical vessel performance dataset 218 includes a historical set of stable vessel operating parameters associated with each vessel of a plurality of vessels. Various examples of the historical set of stable vessel operating parameters of each vessel include but are not limited to at least one of engine power, shaft Revolutions Per Minute (RPM), engine load, engine room temperature, pressure within engine room, power generated by the waste heat recovery system's steam and power turbines, inlet and outlet temperatures of the exhaust gas for turbo charger, temperature and pressure in the scavenging air receiver ofP24-009PCT1main engine and / or auxiliary engine, pressure in the exhaust receiver of main engine and / or auxiliary engine, back pressure of exhaust gas in turbo charger, temperature and viscosity of the fuel oil, average temperature of the exhaust gas, maximum continuous rating, maximum shaft RPM, whether the main engine and / or auxiliary engine has a turbocharger cut-out, number of turbochargers, number of cylinders, diameter of each of the cylinders, number of piston strokes, length of the piston stroke, hull capacity for refrigerated cargo and dimensions, fuel consumption, Speed over ground (SOG), true heading, rudder angle, data on weather conditions, sea state, data on cargo conditions, ballast tank levels, overall fuel levels, fuel type, fuel mix (of mixed fuel batch), hull integrity, machinery condition, distance traveled, estimated time of arrival, actual energy consumption, etc., among other suitable vessel operating parameters.

[0065] In an example, the ML model 220 is an ensemble ML model. Examples of ensemble ML models include, but are not limited to, Random Forest (RF), Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), Adaptive Boosting (AdaBoost), Bootstrap Aggregating (Bagging), Gradient Boosting Machine (GBM), Voting Classifier, Stacked Generalization (Stacking), Multiple Additive Regression Trees (MART), Gradient Boosted Regression Trees (GBRT), and so on. In an implementation, the ML model 220 is an ensemble ML model configured to perform quantile regression with a quantile parameter set to the performance threshold. The training stage of the ML model 220 has been described later with reference to FIG. 4.

[0066] The user interface 212 is an interface such as a Human Machine Interface (HMI) or a software application that allows users such as an administrator (not shown) to interact with and control the server system 200 or one or more parameters associated with the server system 200. It may be noted that the user interface 212 may be composed of several components that vary based on the complexity and purpose of the application. Examples of components of the user interface 212 may include visual elements, controls, navigation, feedback and alerts, user input and interaction, responsive design, user assistance and help, accessibility features, and the like. More specifically these components may correspond to icons, layout, color schemes, buttons, sliders, dropdown menus, tabs, links, error / success messages, mouse and touch interactions, keyboard shortcuts, tooltips, screen readers, and the like.

[0067] The processor 206 includes suitable logic, circuitry, and / or interfaces to execute operations for predicting the energy consumption of the vessel 104, determining a fuel wastage by the vessel 104, and the like. Examples of the processor 206 include, but are not limited to,P24-009PCT1an Application-Specific Integrated Circuit (ASIC) processor, a Reduced Instruction Set Computing (RISC) processor, a Graphical Processing Unit (GPU), a Complex Instruction Set Computing (CISC) processor, a Field-Programmable Gate Array (FPGA), and the like.

[0068] The memory 208 includes suitable logic, circuitry, and / or interfaces to store a set of computer-readable instructions for performing the various operations described herein. Examples of the memory 208 include a random-access memory (RAM), a read-only memory (ROM), a removable storage drive, a hard disk drive (HDD), and the like. It will be apparent to a person skilled in the art that the scope of the disclosure is not limited to realizing the memory 208 in the server system 200, as described herein. In another embodiment, the memory 208 may be realized in the form of a database server or a cloud storage working in conjunction with the server system 200, without departing from the scope of the present disclosure.

[0069] The processor 206 is operatively coupled to the communication interface 210, such that the processor 206 is capable of communicating with a remote device (z.e., to / from a remote device 222) such as third-party servers or with the vessel 104, the one or more data sources 106, or communicating with any entity connected to the network 108 (as shown in FIG. 1).

[0070] It is noted that the server system 200 as illustrated and hereinafter described is merely illustrative of an apparatus that could benefit from embodiments of the present disclosure and, therefore, should not be taken to limit the scope of the present disclosure. It is noted that the server system 200 may include fewer or more components than those depicted in FIG. 2.

[0071] In one implementation, the processor 206 includes a data pre-processing module 224, a data filtering module 226, a consumption prediction module 228, and a graphical user interface (GUI) generation module 230. It should be noted that components, described herein, such as the data pre-processing module 224, the data filtering module 226, the consumption prediction module 228, and the graphical user interface (GUI) generation module 230 can be configured in a variety of ways, including electronic circuitries, digital arithmetic, and logic blocks, and memory systems in combination with software, firmware, and embedded technologies.

[0072] In an embodiment, the data pre-processing module 224 includes suitable logic and / or interfaces for recording the plurality of vessel operating parameters from at least one data source associated with the vessel at one or more frequencies. In particular, the data pre-P24-009PCT1processing module 224 may utilize the one or more data sources 106 to record or access the vessel operating parameters for the vessel 104. As may be understood, vessel operating parameters are dynamic in nature, therefore they have to be recorded at various frequencies. For instance, a few vessel operating parameters have to be recorded at a higher frequency such as every few milliseconds, seconds, minutes, or so on, while others may be recorded at a medium frequency such as every few minutes, hours, and so on, or lower frequency such as every few hours, days, and so on as well. The decision to record different vessel operating parameters at different frequencies may be made based on the type of each vessel operating parameter. For instance, shaft RPM may be recorded at high frequency while weather-related data may be recorded at medium frequency.

[0073] Examples of the vessel operating parameters include, but are not limited to, engine power, shaft Revolutions Per Minute (RPM), engine load, engine room temperature, pressure within engine room, power generated by the waste heat recovery system's steam and power turbines, inlet and outlet temperatures of the exhaust gas for turbo charger, temperature and pressure in the scavenging air receiver of main engine and / or auxiliary engine, pressure in the exhaust receiver of main engine and / or auxiliary engine, back pressure of exhaust gas in turbo charger, temperature and viscosity of the fuel oil, average temperature of the exhaust gas, maximum continuous rating, maximum shaft RPM, whether the main engine and / or auxiliary engine has a turbocharger cut-out, number of turbochargers, number of cylinders, diameter of each of the cylinders, number of piston strokes, length of the piston stroke, hull capacity for refrigerated cargo and dimensions, fuel consumption, Speed over ground (SOG), true heading, rudder angle, data on weather conditions, sea state, data on cargo conditions, ballast tank levels, overall fuel levels, fuel type, fuel mix (of mixed fuel batch), hull integrity, machinery condition, distance traveled, estimated time of arrival, etc., among other suitable vessel operating parameters.

[0074] Further, the data pre-processing module 224 is configured to aggregate the plurality of recorded vessel operating parameters at predefined intervals. This aggregation process generates a set of aggregated vessel operating parameters. As may be appreciated, since vessel operating parameters are recorded at different frequencies (or the same high frequency), there exists a huge amount of values or data that needs to be processed by the server system 200 to obtain an understanding of these parameters. Therefore, by aggregating these parameters over predefined intervals such as 5 min., 10 min., 15 min., and so on, the complexity in understanding these parameters is significantly reduced. Further, fewP24-009PCT1computational resources may be required for analyzing this aggregated data. In an instance, the duration of the predefined interval can be defined by an administrator (not shown) of the server system 200 or an operator of the vessel 104.

[0075] In an embodiment, the data filtering module 226 includes suitable logic and / or interfaces for extracting a set of stable vessel operating parameters from the plurality of aggregated vessel operating parameters based, at least in part, on the stability criteria associated with the vessel. Herein, the stability criteria define a set of stable operating conditions or predefined operating conditions for the vessel 104. In an instance, the stability criteria can be defined by an administrator of the server system 200 or an operator of the vessel 104. The stable vessel operating parameters are selected to ensure the engine consumption (determined later) is determined during stable operating conditions for the vessel 104. It is noted that due to the unpredictable nature of the unstable operating conditions, the energy consumption in such conditions becomes complex to predict. Therefore, to improve the predictive performance of the ML model 220, the predictions are performed using only stable vessel operating parameters. Since, these stable vessel operating parameters are free from disturbance due to temporary transients, weather conditions, etc., among other temporary unstable conditions, they don’t affect the predictive performance of the ML model 220.

[0076] In particular, for extracting the set of stable vessel operating parameters, the data filtering module 226 is configured to identify one or more invalid vessel operating parameters from the plurality of vessel operating parameters based, at least in part, on first filtering criteria within the stability criteria. Herein, the invalid vessel operating parameters indicate that if the one to more stability conditions in the first filtering criteria are not met, then the payload corresponding with the vessel operating parameters would be invalid. The first filtering criteria may be defined based on at least one of a stable engine power range, a stable shaft Revolutions Per Minute (RPM) range, a stable engine load range, a stable fuel consumption range, or a Speed Over Ground (SOG) range. The ‘stable engine power range’ may define an acceptable range of Maximum Continuous Rating (MCR) outside which any value is invalid. The ‘stable shaft RPM range’ may define a stable RPM range outside which any value is invalid. The ‘stable engine load range’ may define a stable range of engine load (for either the main or auxiliary engine) outside which any value is invalid. The ‘stable fuel consumption range’ may define a stable range of fuel consumption outside which any value is invalid. In an instance, the fuel consumption may be in Metric Tonnes per hour (MT / hr) or Kilowatt per hour (kWh). The SOG range may define a stable range SOG GPS outside ofP24-009PCT1which any value is invalid. In an instance, the SOG is measured in Knots.

[0077] Then, the data filtering module 226 is configured to identify one or more frozen vessel operating parameters from the plurality of vessel operating parameters based, at least in part, on second filtering criteria within the stability criteria. Herein, the frozen vessel operating parameters indicate that if the one to more stability conditions in the second filtering criteria are not met, then the sensors associated with the one or more data sources 106 may be frozen.

[0078] The second filtering criteria may be defined based on at least one of a stagnant engine power factor, a stagnant shaft RPM factor, or a stagnant fuel consumption factor. The stagnant engine power factor indicates that if the absolute difference between payloads is zero, then the sensor is frozen and the corresponding value should not be considered. Similarly, the stagnant shaft RPM factor and the stagnant fuel consumption factor respectively indicate that if the absolute difference between payloads is zero, then the sensor is frozen and the corresponding value should not be considered as well.

[0079] Further, the data filtering module 226 eliminates the one or more invalid vessel operating parameters and the one or more frozen vessel operating parameters from the plurality of vessel operating parameters to determine a set of remaining vessel operating parameters. Thereafter, the data filtering module 226 is configured to filter the set of stable vessel operating parameters from the set of remaining vessel operating parameters based, at least in part, on third filtering criteria within the stability criteria. Herein, the stable vessel operating parameters indicate that if the one to more stability conditions in the third filtering criteria are met over a specified time interval (such as 30 min., or so on), then the vessel operating parameters may be called stable.

[0080] The third filtering criteria may be defined based on at least one of a stable engine load range, at least one stable shaft RPM operating range, a stable true heading range, a stable rudder angle range, or a stable SOG operating range. The stable engine load range indicates the difference in the payload for each predefined interval within the specified time interval should be lower than a predefined percentage (such as 5%, 10%, etc. . In an instance, the stable shaft RPM operating range may indicate that the shaft RPM should be greater than a predefined RPM over a rolling window. In another instance, the stable shaft RPM operating range may indicate that the difference between the shaft RPM for each predefined interval should be within the specified time interval and should be lower than another predefined RPM. The stable true heading range may indicate that the difference between the true heading of eachP24-009PCT1predefined interval should be less than a predefined heading angle. The stable rudder angle range may indicate that the difference between the rudder angles of each predefined interval should be less than a predefined rudder angle. The stable SOG operating range may indicate that the SOG should fall within a specific range (e.g., greater than X knots but less than Y knots, herein X and Y are non-zero natural numbers such that X<Y).

[0081] In another embodiment, the data filtering module 226 is configured to determine a subset of stable vessel operating parameters from the set of stable vessel operating parameters. Herein, each stable vessel operating parameter in the subset of stable vessel operating parameters satisfies a performance threshold. In an implementation, the performance threshold is defined by an administrator of the server system 200 or an operator of the vessel 104. For instance, a performance threshold of 10% selects the top 10% of stable vessel operating parameters to form the subset of stable vessel operating parameters. In particular, the performance threshold indicates a selection quantile that has to be accessed or extracted from the subset of stable vessel operating parameters. The selection quantile may be predefined in the database 204 by the administrator. Then, the subset of vessel operating parameters present within the selection quantile is identified and extracted from the set of stable vessel operating parameters.

[0082] In an embodiment, the consumption prediction module 228 includes suitable logic and / or interfaces for predicting an energy consumption for the vessel 104. In various instances, the energy consumption for a specified engine load of the vessel 104 may be predicted in terms of Specific Fuel Oil Consumption (SFOC), Engine Fuel Efficiency (EFE), Engine Power Output (EPO), Total Fuel Consumption (TFC), Energy Consumption per Nautical Mile (kWh / NM), and so on may be predicted to ascertain the energy consumed. In particular, the consumption prediction module 226 is configured to generate a set of features based, at least in part, on the subset of stable vessel operating parameters. It is noted that one or more feature generation techniques (known as feature engineering) such as, but not limited to, time-based feature generation, binning, aggregated variable generation, dimensionality reduction, one hot encoding, frequency encoding, and so on, may be used for generating the features using the subset of stable vessel operating parameters. Since these operations are well known in the art, the same have not been described herein for the sake of brevity.

[0083] Then, the consumption prediction module 228 utilizes the ML model 220 for predicting the energy consumption of the vessel 104 based, at least in part, on applying the set of features on the ML model 220. As may be understood, the energy consumption of the vesselP24-009PCT1104 is predicted for a specific engine load. It is noted that the ML model 220 is trained before its operation during deployment based on the historical vessel performance dataset 218. This aspect has been described later in detail with reference to FIG. 4. It is noted that the energy consumption of the vessel 104 may be computed for specific shipping routes as well. Therefore, by relying on energy consumption predictions for different shipping routes, the operator can determine the best shipping route that will provide the highest fuel efficiency.

[0084] In another embodiment, the consumption prediction module 228 is configured to utilize the data pre-processing module 224 for recording a plurality of actual vessel operating parameters from at least one data source from the one or more data sources 106. It is noted that the actual vessel operating parameters are real-time vessel operating parameters that include the actual fuel consumption of the vessel 104 along with the other vessel operating parameters described earlier. Then, the consumption prediction module 228 is configured to compute an actual energy consumption based, at least in part, on the plurality of actual vessel operating parameters. Further, the consumption prediction module 228 is configured to compare the actual energy consumption with the predicted energy consumption for determining a fuel wastage by the vessel 104. In some instances, the consumption prediction module 228 is configured to compute a fuel consumption value based on the predicted energy consumption of the vessel 104 for a specific shipping route. It is noted that the fuel wastage provides an indication to the operator that there may be a problem with one or more equipment associated with the vessel 104 or the operation of the vessel 104. In response, the operator may take one or more actions to resolve these problems.

[0085] In an embodiment, the GUI generation module 230 includes suitable logic and / or interfaces for generating various GUIs for facilitating the various modules of the server system 200 to perform the various operations described herein. For instance, the GUI generation module 230 may generate a GUI and facilitate a visualization of the comparison between the actual energy consumption with the predicted energy on an electronic device of the operator. In various non-limiting examples, the electronic device may refer to any electronic device such as, but not limited to, Personal Computers (PCs), tablet devices, Personal Digital Assistants (PDAs), Virtual Reality (VR) devices, smartphones, and laptops. The GUI may facilitate the operator in selecting the most fuel-efficient route from a plurality of routes for the vessel 104.

[0086] As may be appreciated, the proposed approach described by the various embodiments herein addresses the various shortcomings such as the lack of validation, limitedP24-009PCT1applicability as well as poor scalability to other vessels of the conventional methods. The proposed approach is scalable to a plurality of vessels (since the ML model 220 is trained on data from different vessels) and provides a stable realistic Main Engine (ME) baseline for the assessment of the maintenance state and proper operation of the main engine by onboard personnel.

[0087] FIG. 3 illustrates a schematic representation of the process 300 for predicting an energy consumption of a vessel 302 such as vessel 104, in accordance with an embodiment of the present disclosure. It is noted that the various operations described in the present disclosure can be divided into two stages, z.e., a training stage and a deployment stage. The training stage of the ML model 220 has been described later with reference to FIG. 4 and the deployment stage is described herein with reference to FIG. 3. It is noted that vessel 302 and one or more data sources 304 of FIG. 3 is identical to vessel 104 and one or more data sources 106 of FIG. 1, respectively. Similarly, ML model 306 of FIG. 3 is identical to ML model 220 of FIG. 2. It is noted that various aspects of FIG.3 have already been explained earlier with reference to FIG. 2 therefore the same are not explained again for the sake of brevity.

[0088] During the deployment stage, the server system 200 is configured to operate at least one data source of one or more data sources 304 to record the vessel operating parameters. Since these vessel operating parameters are often recorded at one or more frequencies, the server system 200 is configured to aggregate these vessel operating parameters at predefined intervals. Later, stable vessel operating parameters are extracted from the aggregated vessel operating parameters. Then, the server system 200 is configured to extract a subset of the stable vessel operating parameters from the stable vessel operating parameters based, at least in part, on the performance threshold. In particular, all stable vessel operating parameters that are at least equal to the performance threshold are added to the subset of the stable vessel operating parameters. Further, the server system 200 is configured to generate a set of features from the subset of stable vessel operating parameters.

[0089] Thereafter, the server system 200 is configured to utilize the ML model 306 for predicting an energy consumption 308 for the vessel 302 based, at least in part, on the set of features. It is noted that the ML model 306 is trained on the historical vessel performance dataset 218 to learn patterns, relationships, and trends in the input data, z.e., the subset of stable vessel operating parameters during deployment. In an implementation, the ML model 306 is a LightGBM model configured to perform quantile regression with a quantile parameter set to the performance threshold. In various examples, the predicted energy consumption 308 mayP24-009PCT1be at least one of an SFOC value, an Engine Fuel Efficiency (EFE) value, an Engine Power Output (EPO) value, a Total Fuel Consumption (TFC) value, an Energy Consumption per Nautical Mile (kWh / NM) value, and / or so on.

[0090] Upon prediction of the energy consumption 308, the server system 200 may be configured to perform one or more additional operations for providing crucial information to the operator of the vessel 302. In one example, the server system 200 is configured to utilize a Graphical User Interface (GUI) 310 for facilitating a visualization of the comparison between the actual energy consumption with the predicted energy consumption on an electronic device associated with the operator. In various non-limiting examples, the electronic device may refer to any electronic device such as, but not limited to, Personal Computers (PCs), tablet devices, Personal Digital Assistants (PDAs), voice-activated assistants, Virtual Reality (VR) devices, smartphones, and laptops. The GUI 310 may facilitate the operator in interacting with the various parameters and results described herein. In some instances, the server system 200 may utilize external services (called via Application Programming Interface (API) for facilitating the generation of the said visualization.

[0091] In another example, the server system 200 is configured to determine a fuel wastage 312 by the vessel 302 during its journey through the planned route. This fuel wastage 312 is determined by comparing the difference between the predicted energy consumption 308 and the actual energy consumption during the said journey. This metric can help the operator to understand, the various instances of fuel wastage 312 during different intervals within the said journey. This information may assist the operator or a route planner in optimizing the route of the vessel 302. Additionally, the server system 200 may perform a comparison between the predicted energy consumption 308 and the actual energy consumption during the said journey. In particular, the various vessel operating parameters may be compared with the subset of stable operating parameters by the server system 200 to determine one or more reasons for the difference between the actual and predicted energy consumption.

[0092] FIG. 4 illustrates a flow diagram of a method 400 of operating the server system 200 for training the Machine Learning (ML) model such as ML model 220 for predicting the energy consumption of a vessel such as the vessel 104, in accordance with an embodiment of the present disclosure. The method 400 depicted in the flow diagram may be executed by, for example, the server system 200. The sequence of operations of the method 400 may not be necessarily executed in the same order as they are presented. Further, one or more operations may be grouped and performed in the form of a single step, or one operation may have severalP24-009PCT1sub-steps that may be performed in parallel or in a sequential manner. Operations of the method 400, and combinations of operations in the method 400 may be implemented by, for example, hardware, firmware, a processor, circuitry, and / or a different device associated with the execution of software that includes one or more computer program instructions. The plurality of operations is depicted in the process flow of the method 400. The process flow starts at operation 402.

[0093] At 402, the method 400 includes accessing a historical vessel performance dataset such as historical vessel performance dataset 218 including a historical set of stable vessel operating parameters associated with each vessel of a plurality of vessels from a database such as database 204. As described earlier, the historical set of stable vessel operating parameters refers to vessel operating parameters that have been historically recorded from one or more data sources (or at least one data source of the one or more data sources) such as one or more data sources 304 associated with different vessels at a one or more frequencies. In other words, these parameters are recorded for each vessel of the plurality of vessels separately and indexed in the historical vessel performance dataset 218. For instance, the historical set of stable vessel operating parameters may be recorded every few milliseconds, seconds, minutes, or so on. In another implementation, the data recording process for a few vessel operating parameters may take place using medium-frequency recording (every few minutes to hours) or low-frequency recording (every few hours or days) as well.

[0094] Various examples of the historical set of stable vessel operating parameters of each vessel include but are not limited to at least one of engine power, shaft Revolutions Per Minute (RPM), engine load, engine room temperature, pressure within engine room, power generated by the waste heat recovery system's steam and power turbines, inlet and outlet temperatures of the exhaust gas for turbo charger, temperature and pressure in the scavenging air receiver of main engine and / or auxiliary engine, pressure in the exhaust receiver of main engine and / or auxiliary engine, back pressure of exhaust gas in turbo charger, temperature and viscosity of the fuel oil, average temperature of the exhaust gas, maximum continuous rating, maximum shaft RPM, whether the main engine and / or auxiliary engine has a turbocharger cut-out, number of turbochargers, number of cylinders, diameter of each of the cylinders, number of piston strokes, length of the piston stroke, hull capacity for refrigerated cargo and dimensions, fuel consumption, Speed over ground (SOG), true heading, rudder angle, data on weather conditions, sea state, data on cargo conditions, ballast tank levels, overall fuel levels, fuel type, fuel mix (of mixed fuel batch), hull integrity, machineryP24-009PCT1condition, distance traveled, estimated time of arrival, actual energy consumption, etc., among other suitable vessel operating parameters. Since these historical vessel operating parameters are recorded at different frequencies (or the same high frequency), these parameters are aggregated at the predefined intervals. Examples of predefined intervals include 5 minutes (min.), 10 min., 15 min., and so on. It is noted that during the model training process, any instances where vessels have operated on a fuel mix (of mixed fuel batch) are filtered out for having a consistent training set.

[0095] At 404, the method 400 includes determining a historical subset of stable vessel operating parameters from the historical set of stable vessel operating parameters for the each vessel. Herein, each historical stable vessel operating parameter in the historical subset of stable vessel operating parameters satisfies the performance threshold. It is noted that the performance threshold may be predefined by an administrator (not shown). In other words, the performance threshold is configurable by an administrator based, at least in part, on different requirements. The performance threshold may define a top threshold percentage for selecting the historical subset of stable vessel operating parameters. The goal of using the performance threshold is to extract the historical vessel operating parameters that are higher than the specified performance threshold. For instance, a performance threshold of 10% selects the top 10% of historical vessel operating parameters to form the historical subset of stable vessel operating parameters. Further, operation 404 is performed for each vessel of the plurality of vessels.

[0096] At 406, the method 400 includes splitting the historical subset of stable vessel operating parameters for the each vessel into a training set and a validation set. As may be understood, the training set is a subset of the data (herein, the historical subset of stable vessel operating parameters) used to train the ML model 220. The training set includes input-output pairs, where the inputs (i.e., features such as the various vessel operating parameters) are used to predict the outputs (i.e., target value such as the energy consumption). The ML model 220 learns from the training set by adjusting its hyperparameters to minimize the prediction error (described later). The goal of using a training set is to enable the ML model 220 to learn patterns, relationships, and trends in the data (i.e., the historical subset of stable vessel operating parameters).

[0097] On the other hand, the validation set is a subset of the data (herein, the historical subset of stable vessel operating parameters) used to evaluate the performance of the ML model 220 during the model training process. The validation set is used to optimize theP24-009PCT1hyperparameters of the ML model 220 for identifying the best model configured to be used during model deployment. In another implementation, the validation set is used to detect overfitting, z.e., identify whether the model performance on new data is good or not.

[0098] In a non-limiting implementation, if the historical subset of stable vessel operating parameters is recorded for a duration of one year, then the training set may include the historical vessel operating parameters for six months, z.e., January to June and the validation set may include the historical vessel operating parameters for the next six months, z.e., July to December. In other instances, the historical subset of stable vessel operating parameters by split such that the training set is 60%, the validation set is 20%, and a test set is 20% of the historical subset of stable vessel operating parameters. Here, the test set is used to test the performance of the trained ML model 220.

[0099] At 408, the method 400 includes iteratively performing the following set of operations, i.e., 408(1) to 408(5) till training criteria are met. The training criteria represent one or more conditions for concluding the model training process. The training criteria include a stage in the iterative process where the change in performance between consecutive iterations of the ML model 220 saturates or improves marginally, the change in loss values (such as quantile loss) computed between consecutive iterations of the ML model 220 saturates or improves marginally, a fixed number of iterations is performed, the training is conducted for a fixed number of epochs, and so on.

[0100] At 408(1), the method 400 includes initializing the ML model 220 based, at least in part, on a plurality of hyperparameters. Herein, the ML model 220 is an ensemble ML model configured to perform quantile regression with a quantile parameter set to the performance threshold. Examples of ensemble ML models include, but are not limited to, Random Forest (RF), Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), Adaptive Boosting (AdaBoost), Bootstrap Aggregating (Bagging), Gradient Boosting Machine (GBM), Voting Classifier, Stacked Generalization (Stacking), Multiple Additive Regression Trees (MART), Gradient Boosted Regression Trees (GBRT), and so on. In a non-limiting implementation, the ML model 220 can be a LightGBM model that is configured to perform a quantile regression process with the quantile parameter set to the performance threshold. As may be understood, when working with real-world regression models, knowing the uncertainty behind each point estimation makes the predictions more actionable in a deployment or real-world setting. One method of going from a single point estimation to a range estimation or so-called prediction interval or the predefined interval isP24-009PCT1known as quantile regression. Quantile regression allows the ML model 220 to establish predictions for an arbitrary quantile, z.e., the performance threshold. The major difference between quantile regression against general regression lies in the loss function, which is called a pinball loss or a quantile loss. Examples of hyperparameters for LightGBM include learning Rate, number of trees, maximum depth, minimum data in leaf, number of leaves, subsample rate, feature fraction, lambda LI, lambda L2, boosting type, early stopping rounds, and so on.

[0101] At 408(2), the method 400 includes generating a set of historical features based, at least in part, on the training set. It is noted that one or more feature generation techniques (known as feature engineering) such as, but not limited to, time-based feature generation, binning, aggregated variable generation, dimensionality reduction, one hot encoding, frequency encoding, and so on, may be used for generating the described features, z.e., the historical features. In some instances, operations that may include removing noise, feature selection, data cleaning, handling missing values, normalizing or scaling data, analyzing characteristics of the data, and converting the data into a format that Al or ML models can process may also be performed. Since these operations are well known in the art, the same have not been described herein for the sake of brevity.

[0102] At 408(3), the method 400 includes generating, by the ML model 220, an energy consumption prediction for at least one time interval associated with the historical subset of stable vessel operating parameters in the validation set, based, at least in part, on applying the set of historical features on the ML model 220. In particular, the ML model 220 utilizes the features generated using the training set to learn the relationship between the features (generated using the various vessel operating features) and the target variable (z.e., the energy consumption prediction).

[0103] At 408(4), the method 400 includes computing, by the ML model 220 using a quantile loss function, at least one quantile loss between the energy consumption prediction and an actual energy consumption computed using the training set and / or the validation set for the at least one time interval. Returning to the previous example, in the quantile regression, the quantile regression loss function may be defined as L, z.e., a tilted or pinball loss function.

[0104] At 408(5), the method 400 includes optimizing the quantile loss function based, at least in part, on the at least one quantile loss to update or tune the plurality of hyperparameters. For optimizing the quantile loss function L, the tilted or pinball loss function takes both ground truth yt(i.e., the actual energy consumption) and approximation yt(i.e., theP24-009PCT1energy consumption prediction or the prediction of the ML model 220) as arguments. Moreover, the specific quantile y [0, 1] is a required input parameter. In a non-liming example, the quantile loss function £ can be defined using Eqn. (1) given below:

[0105] The quantile loss function £ adjusts the weight of the sample’s error according to the given or the targeted quantile y (or the quantile parameter), i.e. a smaller y increases the magnitude of the loss of those samples with a negative residual e.

[0106] Further, the quantile loss function £ introduces more punishment for overestimation. For large y it is vice versa. In a non-limiting scenario, y = 0.1 is used for reflecting baseline conditions. In other words, 10% of the samples fall below, whereas 90% of the data is located above the predicted baseline. It is noted that when the quantile equals 0.5, the loss function is unsymmetrical. For high quantile prediction, the loss function encourages higher prediction value, and vice versa for low quantile prediction. In various examples, the hyperparameters may be tuned using automated hyperparameter tuning techniques such as Optuna, Hyperopt, GridSearchCV, RandomizedSearchCV, BayesianOptimization, Sequential Model-based Algorithm Configuration (SMAC), Keras Tuner, Ray Tune, and so on.

[0107] As may be appreciated, once the ML model 220 is trained, it can be deployed in real-world applications such as predicting energy consumption for a vessel (such as vessel 104), determining fuel wastage using the energy consumption predictions, generating operational recommendations based, at least in part, on the energy consumption predictions, and so on.

[0108] FIG. 5 illustrates a flow diagram of a method 500 of operating the server system 200 for predicting energy consumption for a vessel such as vessel 104, in accordance with an embodiment of the present disclosure. The method 500 depicted in the flow diagram may be executed by, for example, the server system 200. The sequence of operations of the method 500 may not be necessarily executed in the same order as they are presented. Further, one or more operations may be grouped and performed in the form of a single step, or one operation may have several sub-steps that may be performed in parallel or in a sequential manner. Operations of the method 500, and combinations of operations in the method 500 may be implemented by, for example, hardware, firmware, a processor, circuitry, and / or a different device associated with the execution of software that includes one or more computer programP24-009PCT1instructions. The plurality of operations is depicted in the process flow of the method 500. The process flow starts at operation 502.

[0109] At 502, the method 500 includes accessing a set of stable vessel operating parameters from a plurality of vessel operating parameters of a vessel such as vessel 104, recorded at predefined intervals for the vessel 104. Herein, the set of stable vessel operating parameters satisfies stability criteria. As described earlier, the stability criteria are predefined by an administrator (not shown) associated with the server system 200. In an instance, the stability criteria include a set of predefined operating conditions for the vessel 104.

[0110] At 504, the method 500 includes determining a subset of stable vessel operating parameters from the set of stable vessel operating parameters. Herein, each stable vessel operating parameter in the subset of stable vessel operating parameters satisfies a performance threshold. As described earlier, the performance threshold is predefined by the administrator associated with the server system 200. For example, the performance threshold may be set to the top 10% to extract the top 10% of the vessel operating parameters within the set of stable vessel operating parameters to form the subset of stable vessel operating parameters.

[0111] At 506, the method 500 includes generating a set of features based, at least in part, on the subset of stable vessel operating parameters.

[0112] At 508, the method 500 includes predicting, by a Machine Learning (ML) model such as ML model 220, an energy consumption for the vessel 104 based, at least in part, on applying the set of features on the ML model 220.

[0113] The disclosed method with reference to FIG. 4 and FIG. 5, or one or more operations of the server system 200 may be implemented using software including computerexecutable instructions stored on one or more computer-readable media (e.g., non-transitory computer-readable media, such as one or more optical media discs, volatile memory components (e.g., DRAM or SRAM), or nonvolatile memory or storage components (e.g., hard drives or solid-state nonvolatile memory components, such as Flash memory components) and executed on a computer (c.g, any suitable computer, such as a laptop computer, netbook, Web book, tablet computing device, smartphone, or other mobile computing devices). Such software may be executed, for example, on a single local computer or in a network environment (e.g., via the Internet, a wide-area network, a local-area network, a remote web-based server, a client-server network (such as a cloud computing network), or other such networks) using one or more network computers.P24-009PCT1

[0114] Additionally, any of the intermediate or final data created and used during the implementation of the disclosed methods or systems may also be stored on one or more computer-readable media (e.g., non-transitory computer-readable media) and are considered to be within the scope of the disclosed technology. Furthermore, any of the software-based embodiments may be uploaded, downloaded, or remotely accessed through a suitable communication means. Such suitable communication means include, for example, the Internet, the World Wide Web (WWW), an intranet, software applications, cable (including fiber optic cable), magnetic communications, electromagnetic communications (including RF, microwave, and infrared communications), electronic communications, or other such communication means.

[0115] Although the invention has been described with reference to specific exemplary embodiments, it is noted that various modifications and changes may be made to these embodiments without departing from the broad scope of the invention. For example, the various operations, blocks, etc., described herein may be enabled and operated using hardware circuitry (for example, Complementary Metal Oxide Semiconductor (CMOS) based logic circuitry), firmware, software, and / or any combination of hardware, firmware, and / or software (for example, embodied in a machine-readable medium). For example, the apparatuses and methods may be embodied using transistors, logic gates, and electrical circuits (for example, Application Specific Integrated Circuit (ASIC) circuitry and / or Digital Signal Processor (DSP) circuitry).

[0116] Particularly, the server system 200 and its various components may be enabled using software and / or using transistors, logic gates, and electrical circuits (for example, integrated circuit circuitry such as ASIC circuitry). Various embodiments of the invention may include one or more computer programs stored or otherwise embodied on a computer-readable medium, wherein the computer programs are configured to cause the processor or the computer to perform one or more operations. A computer-readable medium storing, embodying, or encoded with a computer program, or similar language, may be embodied as a tangible data storage device storing one or more software programs that are configured to cause the processor or computer to perform one or more operations. Such operations may be, for example, any of the steps or operations described herein. In some embodiments, the computer programs may be stored and provided to a computer using any type of non-transitory computer- readable media. Non-transitory computer-readable media includes any type of tangible storage media.P24-009PCT1

[0117] Examples of non-transitory computer-readable media include magnetic storage media (such as floppy disks, magnetic tapes, hard disk drives, etc. , optical magnetic storage media (e.g. magneto-optical disks), Compact Disc Read-Only Memory (CD-ROM), Compact Disc Recordable (CD-R), compact disc rewritable (CD-R / W), Digital Versatile Disc (DVD), BLU-RAY® Disc (BD), and semiconductor memories (such as mask ROM, programmable ROM (PROM), (erasable PROM), flash memory, Random Access Memory (RAM), etc. . Additionally, a tangible data storage device may be embodied as one or more volatile memory devices, one or more non-volatile memory devices, and / or a combination of one or more volatile memory devices and non-volatile memory devices. In some embodiments, the computer programs may be provided to a computer using any type of transitory computer- readable media. Examples of transitory computer-readable media include electric signals, optical signals, and electromagnetic waves. Transitory computer-readable media can provide the program to a computer via a wired communication line (e.g., electric wires, and optical fibers) or a wireless communication line.

[0118] Various embodiments of the invention, as discussed above, may be practiced with steps and / or operations in a different order, and / or with hardware elements in configurations, which are different than those which, are disclosed. Therefore, although the invention has been described based on these exemplary embodiments, it is noted that certain modifications, variations, and alternative constructions may be apparent and well within the scope of the invention.

[0119] Although various exemplary embodiments of the invention are described herein in a language specific to structural features and / or methodological acts, the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as exemplary forms of implementing the claims.P24-009PCT1

Claims

1. 33CLAIMS1. A computer-implemented method, comprising: accessing a set of stable vessel operating parameters from a plurality of vessel operating parameters of a vessel, recorded at predefined intervals for the vessel, wherein the set of stable vessel operating parameters satisfies stability criteria; determining a subset of stable vessel operating parameters from the set of stable vessel operating parameters, wherein each stable vessel operating parameter in the subset of stable vessel operating parameters satisfies a performance threshold; generating a set of features based, at least in part, on the subset of stable vessel operating parameters; and predicting, by a Machine Learning (ML) model, an energy consumption for the vessel based, at least in part, on applying the set of features on the ML model.

2. The computer-implemented method as claimed in claim 1, further comprising: accessing a historical vessel performance dataset comprising a historical set of stable vessel operating parameters associated with each vessel of a plurality of vessels from a database; determining a historical subset of stable vessel operating parameters from the historical set of stable vessel operating parameters for the each vessel, wherein each historical stable vessel operating parameter in the historical subset of stable vessel operating parameters satisfies the performance threshold; and training, the ML model based, at least in part, on the historical subset of stable vessel operating parameters.

3. The computer-implemented method as claimed in 2, wherein training the ML model comprises: splitting the historical subset of stable vessel operating parameters for the each vessel into a training set and a validation set; and iteratively performing following set of operations till training criteria are met: initializing the ML model based, at least in part, on a plurality of hyperparameters, wherein the ML model is an ensemble ML model configured to perform quantile regression with a quantile parameter set to the performance threshold; generating a set of historical features based, at least in part, on the training set;P24-009PCT134 generating, by the ML model, an energy consumption prediction for at least one time interval associated with the historical subset of stable vessel operating parameters in the validation set, based on applying the set of historical features on the ML model; computing, by the ML model using a quantile loss function, at least one quantile loss between the energy consumption prediction and an actual energy consumption computed using the validation set for the at least one time interval; and optimizing the quantile loss function based on the at least one quantile loss to update the plurality of hyperparameters.

4. The computer-implemented method as claimed in claim 1, wherein accessing the set of stable vessel operating parameters comprises: recording the plurality of vessel operating parameters from at least one data source associated with the vessel at one or more frequencies; aggregating the plurality of recorded vessel operating parameters at the predefined intervals; and extracting the set of stable vessel operating parameters from the plurality of aggregated vessel operating parameters based, at least in part, on the stability criteria associated with the vessel, wherein the stability criteria define stable operating conditions for the vessel.

5. The computer-implemented method as claimed in claim 4, wherein extracting the set of stable vessel operating parameters comprises: identifying one or more invalid vessel operating parameters from the plurality of vessel operating parameters based, at least in part, on first filtering criteria within the stability criteria, wherein the first filtering criteria is based on at least one of a stable engine power range, a stable shaft Revolutions Per Minute (RPM) range, a stable engine load range, a stable fuel consumption range, or a Speed Over Ground (SOG) range; identifying one or more frozen vessel operating parameters from the plurality of vessel operating parameters based, at least in part, on second filtering criteria within the stability criteria, wherein the second filtering criteria is based on at least one of a stagnant engine power factor, a stagnant shaft RPM factor, or a stagnant fuel consumption factor; and eliminating the one or more invalid vessel operating parameters and the one or more frozen vessel operating parameters from the plurality of vessel operating parameters to determine a set of remaining vessel operating parameters.P24-009PCT16. The computer-implemented method as claimed in claim 5, wherein extracting the set of stable vessel operating parameters further comprises: filtering the set of stable vessel operating parameters from the set of remaining vessel operating parameters based, at least in part, on third filtering criteria within the stability criteria, wherein the third filtering criteria is based on at least one of a stable engine load range, at least one stable shaft RPM operating range, a stable true heading range, a stable rudder angle range, or a stable SOG operating range.

7. The computer-implemented method as claimed in claim 1, wherein determining the subset of stable vessel operating parameters comprises: accessing the performance threshold indicating a selection quantile; and identifying and extracting the subset of vessel operating parameters present within the selection quantile from the set of stable vessel operating parameters.

8. The computer-implemented method as claimed in claim 1, further comprising: recording a plurality of actual vessel operating parameters from at least one data source; computing an actual energy consumption based, at least in part, on the plurality of actual vessel operating parameters; comparing the actual energy consumption with the predicted energy consumption; and determining a fuel wastage by the vessel based, at least in part, on the comparison step.

9. The computer-implemented method as claimed in claims 8, further comprising: facilitating generation of at least one Graphical User Interface (GUI) based, at least in part, on the comparing step, wherein the at least one GUI provides a visualization of the comparison between the actual energy consumption with the predicted energy consumption.

10. A server system, comprising: a communication interface; a memory configured to store instructions; and a processor in communication with the communication interface and the memory, the processor configured to execute the instructions stored in the memory and thereby cause the server system to perform at least in part to: access a set of stable vessel operating parameters from a plurality of vessel operating parameters of a vessel, recorded at predefined intervals for the vessel, whereinP24-009PCT1the set of stable vessel operating parameters satisfies stability criteria; determine a subset of stable vessel operating parameters from the set of stable vessel operating parameters, wherein each stable vessel operating parameter in the subset of stable vessel operating parameters satisfies a performance threshold; generate a set of features based, at least in part, on the subset of stable vessel operating parameters; and predict, by a Machine Learning (ML) model, an energy for the vessel based, at least in part, on applying the set of features on the ML model.

11. The server system as claimed in claim 11, wherein the server system is further caused, at least in part, to: access a historical vessel performance dataset comprising a historical set of stable vessel operating parameters associated with each vessel of a plurality of vessels from a database; determine a historical subset of stable vessel operating parameters from the historical set of stable vessel operating parameters for the each vessel, wherein each historical stable vessel operating parameter in the historical subset of stable vessel operating parameters satisfies the performance threshold; and train the ML model based, at least in part, on the historical subset of stable vessel operating parameters.

12. The server system as claimed in claim 12, wherein to train the ML model, the server system is caused, at least in part, to: split the historical subset of stable vessel operating parameters for the each vessel into a training set and a validation set; and iteratively perform following set of operations till training criteria are met: initializing the ML model based, at least in part, on a plurality of hyperparameters, wherein the ML model is an ensemble ML model configured to perform quantile regression with a quantile parameter set to the performance threshold; generating a set of historical features based, at least in part, on the training set; generating, by the ML model, an energy consumption prediction for at least one time interval associated with the historical subset of stable vessel operating parameters in the validation set, based on applying the set of historical features on the ML model; computing, by the ML model using a quantile loss function, at least one quantileP24-009PCT137 loss between the energy consumption prediction and an actual energy consumption computed using the validation set for the at least one time interval; and optimizing the quantile loss function based on the at least one quantile loss to update the plurality of hyperparameters.

13. The server system as claimed in claim 11, wherein to determine the subset of stable vessel operating parameters, the server system is caused, at least in part, to: access the performance threshold indicating a selection quantile; and identify and extract the subset of vessel operating parameters present within the selection quantile from the set of stable vessel operating parameters.

14. The server system as claimed in claim 11, wherein the server system is further caused, at least in part, to: record a plurality of actual vessel operating parameters from at least one data source; compute an actual energy consumption based, at least in part, on the plurality of actual vessel operating parameters; and compare the actual energy consumption with the predicted energy consumption; and determine a fuel wastage by the vessel based, at least in part, on the comparison step.

15. A non-transitory computer-readable storage medium comprising computer-executable instructions that, when executed by at least a processor of a server system, cause the server system to perform a method comprising: accessing a set of stable vessel operating parameters from a plurality of vessel operating parameters of a vessel, recorded at predefined intervals for the vessel, wherein the set of stable vessel operating parameters satisfies stability criteria; determining a subset of stable vessel operating parameters from the set of stable vessel operating parameters, wherein each stable vessel operating parameter in the subset of stable vessel operating parameters satisfies a performance threshold; generating a set of features based, at least in part, on the subset of stable vessel operating parameters; and predicting, by a Machine Learning (ML) model, an energy consumption for the vessel based, at least in part, on applying the set of features on the ML model.P24-009PCT1

Images