System and method to facilitate dynamic storage and retrieval prioritization in a distribution center and / or supply chain

Abstract

A system (100) for a supply chain ecosystem comprising computing devices (160, 260, 460) having a processor (264) and memory storing computer-executable instructions that, when executed by the processor (264), cause the system (100) to obtain extrinsic data (1102) for processing. During a first processing, the extrinsic data (1102) is processed using a first neural network model (266, 466) trained with extrinsic training data to generate a first set of processed data (1104). Intrinsic data (1202) is obtained for processing. During a second processing, the intrinsic data (1202) and the first set of processed data (1104) are processed using a second neural network model (266, 466) trained with intrinsic training data and processed extrinsic training data to generate a second set of processed data (1204). Based on the second set of processed data (1204n), one or more unplanned and unanticipated events to a supply chain disruption are determined.

Description

SYSTEM AND METHOD TO FACILITATE DYNAMIC STORAGE AND RETRIEVAL PRIORITIZATION IN A DISTRIBUTION CENTER AND / OR SUPPLY CHAIN CROSS REFERENCE TO RELATED APPLICATION

[0001] The present application claims the priority benefits of U.S. provisional patent application Serial No. 63 / 615,367, filed on December 28, 2023.FIELD OF THE INVENTION

[0002] The present invention is directed to network systems, and in particular to operational systems responsible for procurement, manufacturing, and distribution activities in a distribution center and / or supply chain.BACKGROUND OF THE INVENTION

[0003] A supply chain is a networked platform of different entities responsible for procurement, manufacturing, warehousing, and distribution activities involved in moving products and / or services from companies to customers.

[0004] In recent years, supply chains have operated in increasingly complex and uncertain environments that are prone to disruptions from both internal and external forces that can be difficult to predict, as well as laborious process given the large number of interdependencies the various entities involved. The surge in big data the supply chains need to digest and the increasing demand for business analytics are experiencing a significant boost. However, reliable forecasts based on the analyses of large datasets represents a fundamental challenge. Traditional machine learning and even shallow neural networks are no longer able to analyze all available data or potential disruptions for resolution within a reasonable amount of time. Examples of such internal and external disruptions include weather, traffic, failure of an automation device / machinery, shortage of inventory relative to product demand, under-resource, as well as others. As an example, producers may maintain stockpiles of articles such as source materials or products produced from those source materials to ensure delivery to customers with minimal delays and maintain business continuity to mitigate these disruptions. However, increases in stockpiles not only tie up capital but also require increases in storage areas. In some cases, a warehouse for stocking articles may be equipped with human-driven trucks, autonomous guided vehicles, autonomous mobile vehicles, drones, or the like for lifting, moving, picking, loading, unloading articles forcoordination of at least one of replenished of articles from another entities, freight and shipment transactions, and so forth, is thus unable to make room to accommodate the increase in stockpiles within a reasonable time.SUMMARY OF THE INVENTION

[0005] Embodiments of the disclosure disclose a computer-implemented system and method for a supply chain ecosystem. The system may include one or more computing devices having a processor and a memory storing computer-executable instructions that, when executed by the processor, cause the system to obtain extrinsic data for training, perform a first training using a first neural network model with the extrinsic data, obtain intrinsic data for training, perform a second training using a second neural network model with the first trained and the intrinsic data, and based on the second trained data, determine one or more of unplanned and unanticipated events to a supply chain disruption. An alert message of the unplanned and unanticipated events is generated and may include one or more recommendations for the supply chain disruption. For instance, the recommendations may include warehouse control operation, machine control operation, robotic control operation, vehicle control operation, stock or order replenishment, storage space prioritizing operation, freight operation, transportation, maintenance operation, order rescheduling, or labor resource planning. The first and second trainings may be training routines that are performed and / or conducted using the noted data.

[0006] In various embodiments, the first and second neural network model may include Artificial Neural Network (ANN), Quantum Neural Network, Reinforced Neural Network, Deep Neural Network (DNN), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Large Language Model (LLM), Multilayer Perceptron (MLP), Generative Adversarial Network (GAN), Fuzzy Neural Network (FNN), Spiking Neural Network (SNN), and combination thereof. In an alternative embodiment, the system is a system on chip (SoC) having the processor and memory.

[0007] The one or more computing devices comprise an intelligent network communicatively coupled to at least one of the processor, and the memory is formed either as a part of a system-on-chip (SoC) or as a part of a system-in-package (SiP)configuration. Each of the extrinsic data and intrinsic data include information in various formats such as text, image, video, meta-data, page links, tweets, author details, data, GPS coordinates, comments, CSS, or combination thereof. The system may further include a control tower and the intelligent network formed as part of the control tower.

[0008] In accordance with another embodiment, a non-transitory computer-readable medium for tracking at least one or more of external forces or activities such as weather and human induced activities for a pre-inbound operations or an incoming operations is disclosed.

[0009] In accordance with yet another embodiment, a non-transitory computer-readable medium having stored thereon executable instructions that, in response to execution, cause one or more computing devices to perform operations is disclosed. The operations may include receiving a first array of aggregated attribute data, performing a first training of an artificial learning model with the first array of aggregated attribute data, identifying data corresponding to a type of condition. The data includes information learned by the artificial learning model. The operations further include receiving a second array of aggregated attribute data associated with the identified data that include information learned by the artificial learning model, performing a second training of an artificial learning model with both the identified data learned by the artificial learning model and the second array of aggregated attribute data, defines as an enhanced data, classifying a grade for the enhanced data, determining whether the grade for the enhanced data is below, within, or above a predetermined threshold to a disruption and transmitting an alert message corresponding to the disruption if the data is graded above a predetermined threshold. The alert message includes one or more recommendations for responding to the disruption. The computing devices may be at least one of a cloud-based artificial computing device or a virtual computing device and may include a processor, a memory, a data store, and a control towel.

[0010] A further embodiment of the disclosure discloses a system and method for training an intelligent system for a supply chain ecosystem. The system may process first and second datasets with a first neural network of the intelligent system using adeep learning model, output a first trained output learned from first and second datasets, add the first trained output into an input layer of a second neural network of the intelligent system, process a third dataset and the first trained output with the second neural network using a feature learning model, and output a second trained output, wherein the second trained output includes (i) at least one unplanned and unanticipated event and (ii) a solution to the unplanned and unanticipated event. The data corresponding to the unplanned and unanticipated event may include a value either above, below, or within a predetermined threshold. If the value for the unplanned and unanticipated event is above the predetermined threshold, the system generates an alert message for the solution to the unplanned and unanticipated event.

[0011] In some cases, the first and second datasets correspond to a first time interval at a first location whereas the first trained output and the third dataset correspond to a second time interval at a second location.

[0012] An embodiment to a system-in-package (SiP) for supply chain disruption and configured to support a plurality of data types is disclosed. The system-on-package includes a processor system having first and second processors and a non-transitory computer-readable medium on which is stored instructions for execution by the processor system. The SiP is configured to process first and second datasets with the first processor using a deep learning model, output a first trained output data learned from first and second datasets, add the first trained output into the second processor, process a third dataset and the first trained output data with the second processor using a feature learning model, and output a second trained output data. The second trained output data may include at least one unplanned and unanticipated event as well as a solution to the unplanned and unanticipated event. The SiP may further determine if data corresponds to the unplanned and unanticipated event includes a value either above, below, or within a predetermined threshold. If the value for the unplanned and unanticipated event is above the predetermined threshold, the SiP generates an alert message for the solution to the unplanned and unanticipated event.

[0013] A further alternative embodiment to a distributed intelligent system for a supply chain ecosystem comprising a plurality of facility entities is disclosed. The system includes a master intelligent network, a plurality of slave intelligent networks, and a provider network. The slave intelligent networks are each communicatively coupled to at least one respective distribution network of the plurality of distribution networks. The provider network is communicatively coupled to the master intelligent network and provides a first set of extrinsic data to the master intelligent network. The master intelligent network comprises a computer with a master neural network model operable to process the first set of extrinsic data to output a first set of processed data. The master intelligent network selects respective subsets of processed data from the first set of processed data to provide to selected ones of the plurality of slave intelligent networks. A first slave intelligent network of the plurality of slave intelligent networks includes a computer and a second slave neural network model for receiving and processing a respective first subset of processed data and a first set of intrinsic data related to a corresponding first distribution network to provide a first set of enhanced data. The first set of enhanced data includes at least one unplanned and unanticipated event affecting the first distribution network and a solution to the unplanned and unanticipated event

[0014] In a further embodiment, the system will transmit an alert message of the unplanned and unanticipated events and will generate one or more recommendations for responding to the supply chain disruption. Such recommendations may include one or more recommendations selected from warehouse control operation, machine control operation, robotic control operation, vehicle control operation, stock or order replenishment, storage space prioritizing operation, freight operation, transportation, maintenance operation, order rescheduling, and labor resource planning.

[0015] In a further embodiment, the extrinsic data and the extrinsic training data are generated by a provider network, and the intrinsic data and the intrinsic training data are generated by one or more facility entities of the supply chain ecosystem. In an alternative embodiment, the provider network is one or more of the following: a weather station, a traffic station, a news station, and a social media outlet; while the intrinsic datais generated by a supply chain ecosystem having one or more facility entities selected from: a supplier, a manufacturer, a distribution center, a warehouse, and an e- commerce. The system may also extract information from the extrinsic data and the intrinsic data during a first processing and / or a second processing. The information extracted from the extrinsic data and the intrinsic data may include one or more datasets selected from a group consisting of: text, images, video, metadata, page links, tweets, author details, data, GPS coordinates, comments, and CSS. The extrinsic data and the extrinsic training data may include one or more selections from the following: weather data, traffic data, news data, data from social media feeds, and event data.The intrinsic data and the intrinsic training data may include one or more selections from the following: unloading schedule, reserve / storage schedule, sortation / packing schedule, order picking schedule, loading status, robots schedule, and vehicle schedule. In an alternative embodiment, the extrinsic data and the intrinsic data are live data. The extrinsic training data and the intrinsic training data are one of live data, previously recorded data, or a combination of live data and previously recorded data.

[0016] In an alternative embodiment, the one or more computing devices include at least one cloud-based artificial computing device or a virtual computing device.

[0017] In yet another alternative embodiment, the computer of the master intelligent network trains the master neural network model to process the extrinsic data, such that the processed data includes a relevant portion of the extrinsic data corresponding to an unplanned and unanticipated event.

[0018] In another alternative embodiment, the master intelligent network trains a first slave neural network model and a second slave neural network model and provides them to a selected slave intelligent network of the plurality of slave intelligent networks. In a further alternative embodiment, the computer of the first slave intelligent network forwards a trained first slave neural network model and a trained second slave neural network model to a second slave intelligent network of the plurality of slave intelligent networks.

[0019] In some cases, the first and second datasets correspond to first time interval at first location. In other cases, the first trained output data and the third dataset correspond to second time interval at second location.BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Below, exemplary embodiments of the disclosure are described in detail with reference to the figures.

[0021] FIG. 1 is a simplified schematic diagram of an intelligent supply chain ecosystem according to an exemplary embodiment of the disclosure;

[0022] FIG. 2 is a simplified schematic diagram of an intelligent supply chain ecosystem according to another exemplary embodiment of the disclosure;

[0023] FIG. 3 is a simplified schematic diagram of a cloud-based control tower for a supply chain ecosystem according to another exemplary embodiment of the disclosure;

[0024] FIG. 4 is a schematic diagram of a distributed computing environment for a supply chain ecosystem according to another exemplary embodiment of the disclosure;

[0025] FIG. 5 is a block diagram of an intelligent supply chain ecosystem having cooperating modules according to another exemplary embodiment of the disclosure;

[0026] FIG. 6 is a block diagram of an intelligent supply chain ecosystem having cooperating modules according to another exemplary embodiment of the disclosure;

[0027] FIG. 7 illustrates example of a deep neural network (DNN) according to an exemplary embodiment of the disclosure;

[0028] FIG. 8 illustrates example of a deep neural network (DNN) incorporating reinforcement learning (RL) model according to an exemplary embodiment of the disclosure;

[0029] FIG. 9A is a simplified block diagram of a weight pattern of 3D extrinsic data or external forces / activities according to an exemplary embodiment of the disclosure;

[0030] FIG. 9B is a simplified block diagram of a weight pattern of 3D intrinsic data according to an exemplary embodiment of the disclosure;

[0031] FIG. 10A is a block diagram of first and second neural networks stored on an intelligent network for a supply chain ecosystem according to an exemplary embodiment of the disclosure; and

[0032] FIG. 10B is another block diagram of first and second neural networks stored on an intelligent network for a supply chain ecosystem according to an exemplary embodiment of the disclosure.DETAILED DESCRIPTION

[0033] For the purposes of promoting an understanding of the principles of the disclosure, reference will now be made to the embodiments illustrated in the drawings and described in the following written specification, wherein like reference numerals represent like parts and assemblies throughout the several views. It is understood that no limitation to the scope of the disclosure is thereby intended. It is further understood that the disclosure includes any alternations and modifications to the illustrated embodiments and includes further applications of the principles of the disclosure as would normally occur to a person of ordinary skill in the art to which this disclosure pertains.

[0034] A supply chain is a networked platform of different entities responsible for procurement, manufacturing, warehousing, and distribution activities involved in moving products and / or services from companies to customers. A supply chain disruption can be defined as unplanned and unanticipated events that disrupt a normal flow of goods and materials within the supply chain and, as consequence, expose facility entities within the supply chain to risks such as operational and management risks. For instance, machine downtime, inventory shortage, labor storage, are some examples of the supply chain disruption.

[0035] Embodiments of the disclosure disclose a computer-implemented system, a computer-implemented method, and a non-transitory computer-readable medium for tracking at least one or more of external forces or activities, such as weather and human induced activities for a pre-inbound operations or an incoming operations. In some cases, the external forces and / or human induced activities used as training data may be in its raw form. This data is transmitted to an intelligent network for training using the artificial learning model. In some cases, the raw data can be smoothed,truncated, aggregated, clustered, or otherwise manipulated into another form, which can then be used for training the same or different artificial learning model. If the artificial learning model used for training the intelligent network is also responsible for predictive disruption, the training data is labeled and classified into several levels or groups such as, for example: no delay, reasonable delay within a predetermined threshold, and an extreme delay beyond the predetermined threshold. For example, the artificial learning model is capable of being trained to analyze the data and providing a result that includes a classification of the datasets into a particular class, group, or level that may cause change in a warehouse inventory management operation. In one embodiment, a warehouse inventory management system for managing / controlling the warehouse inventory management operation can be incorporated into such as a control tower. In another embodiment, the warehouse inventory management system may be communicatively coupled to the control tower either remotely or locally. The artificial learning model can be continuously enhanced and take into account the impact of the changes in the warehouse operations triggered by external impacts / forces. If an output or the result corresponds to the delay exceeding the predetermined threshold or expected reasonable timeline, the artificial learning model may infer a disruption to a particular shipment. A WMS based on output from the artificial learning model of the intelligent network can responsively order for example, a receiving / unloading area affected by the disruption or impact of the changes to open up, unreserve, prioritize space at the receiving / unloading area to accommodate delay in the particular shipment. The artificial learning model can also send a message in various formats, such as text, audio, video, or combination thereof alerting the distribution center affected by the disruption to take one or more corrective measures. For instance, the message is sent to at least one or more of the robots and vehicles to either clear or unreserve the storage space on storage racks in order to accommodate the delay in shipment. The WMS continues to monitor the receiving pipeline and clear the next shipment. The WMS reserves the space on the racks for the next shipment and sends message to the other processing outbound units, going into transportation, logistics and yard to prepare for the next priority wave of shipment. When the disruption is restored to normal, and delayed shipment is cleared, the WMS reassesses the priority dynamically.

[0036] Referring to FIG. 1 , the numeral 100 generally designates an exemplary intelligent supply chain ecosystem, in accordance with an embodiment. The intelligent supply chain ecosystem 100 includes a plurality of facility nodes or entities 102, 104, 106, 108, 110, 112 communicatively coupled to an intelligent network 160 indirectly or directly via communication links L1 -L5. Examples of such facility nodes include, but are not limited to, manufacturers 102, distribution centers 104, transportation 106, shipping 108, e-commerce / grocery 110, consumer 112, warehousing (not shown), wholesalers (not shown), suppliers, and so forth. Although one intelligent supply chain ecosystem 100 is illustrated, embodiments contemplate any number of supply chain ecosystem 100, according to particular needs. Each facility node can perform one or more specific roles / tasks within the supply chain ecosystem 100. In some cases, a single facility node may perform multiple roles / tasks in addition to its own specific roles / tasks within the supply chain ecosystem 100 without departing from the scope of the disclosure. As illustrated, the facility nodes 102, 104, 106, 108, 110, 112 are located in the same region. In some cases, the facility nodes 102, 104, 106, 108, 110, 112 can be located in various geographical locations / areas interconnected over the intelligent network 160.

[0037] The supply chain ecosystem 100 further includes a provider network 140 communicatively coupled directly or indirectly to the intelligent network 160 via a communication link L6. The provider network 140 includes a plurality of service providers, such as a weather center 140a, a transportation center 140b, a News station 140c, a social media center 140d, an event station (not shown), and so forth, configured to gather an array of extrinsic data induced or generated by the one or more of the service providers. For example, the extrinsic data can include weather data, traffic data, date from news feeds, data from social media feeds, social unrest data, environment occurrence and natural disaster data (such as lockdown, pandemic, war, earthquakes, wildfires, cyclones, typhoons, tornadoes, floods), or other suitable data. As illustrated, the plurality of service providers can be located in the same region that operate together as a single provider network 140. In some cases, a plurality of service providers that form multiple provider networks may be located in various geographical locations / areas.

[0038] The extrinsic data may be induced or generated in real-time, either stored as historic data or forwarded as real-time / current data to the supply chain ecosystem 100 over the intelligent network 160. The extrinsic data either in real-time, stored data, or combination thereof may be further trained and inferenced to determine unplanned and unanticipated events that may cause a disruption within the supply chain. A computer- implemented training method, a computer-implemented inference method, a training system, an inference system, or a combination thereof for training and inferring first and second neural networks, such as deep neural networks, for use as a deep training model and inference model to determine unplanned and unanticipated events that may cause a disruption within the supply chain may be provided on the intelligent network 160, as depicted and described in further detail with respect to FIG. 2. In some cases, the computer-implemented training method, the computer-implemented inference method, the training system, the inference system, or combination thereof for training and inferring first and second neural networks, such as deep neural networks, for use as a deep training model and inference model can be implemented on a component such as a microprocessor, a computer processing unit, a graphical processing unit, a silicon-on-chip, a graphene-on-chip, a neural network-on-chip, a neuromorphic chip (NeuRRAM), a quantum computer performing quantum computing, a system on chip (SoC), a system-in-package (SIP) configuration, any suitable component, or any suitable combination of components, as depicted and described in further detail with respect to foregoing figures.

[0039] A user equipment UE may be communicatively coupled directly or indirectly to the supply chain ecosystem, the provider network 140, and the intelligent network 160 via a communication link L7. The user equipment UE may include, for example, a portable computer, a tablet, a laptop computer, a personal computer, a personal digital assistant (PDA), a workstation, a web-browsing device, a smart phone, a smart wearable device (such as a watch, a ring, a bracelet, a goggle, an eyeglasses, or the like), a gaming console, or any suitable computational and / or communicational device having communication capability. The user equipment UE can retrieve, query, track / monitor, archive / store, and / or display information associated with the extrinsic data, the intrinsic data, and / or data being trained and / or inferenced in various formats,such as text, graph, image, audio, video, to a user. The user equipment UE can also forward the information to one or more recipients / facility entities regarding a supply chain disruption, a recommendation, an action, and / or a solution to the supply chain disruption.

[0040] The communication links L1 -L7 may be wired, wireless, or any link suitable to support data communications and data packets between supply chain ecosystem 100, the provider network 140, the intelligent network 160, and the user equipment UE. The intelligent network 160 may be referred as an intelligent module, an intelligent component, an intelligent layer, a training model, an inference model, or any suitable intelligent network that includes a computer-implemented training method, a computer- implemented inference method, a training system, an inference system, or combination thereof for training and inferring first and second neural networks, such as deep neural networks, to determine unplanned and unanticipated events that may cause a disruption within a supply chain ecosystem 100. In some cases, the intelligent network 160 is formed as a subset, a portion, or a sublayer of the intelligent network 160. The intelligent network 160 may be integrated into at least one or more of control tower, Warehouse Management System (WMS), Warehouse Execution System (WES), Warehouse Control System (WCS), or any suitable warehouse operating system, as depicted and described in further detail with respect to FIG. 3.

[0041] In another embodiment, the intelligent network 160 is a standalone network communicatively coupled directly or indirectly to at least one or more of a control tower, a Warehouse Management System (WMS), a Warehouse Execution System (WES), a Warehouse Control System (WCS), or any suitable warehouse operating system, as depicted and described in further detail with respect to FIG. 2.

[0042] In yet another embodiment, each supply chain ecosystem includes a slave intelligent network, and the one or more slave intelligent networks may be communicatively coupled to the provider network 140, with the one or more supply ecosystems coupled to a master intelligent network via communication links, as depicted and described in further detail with respect to FIG. 4. In some cases, acombination of the intelligent network 160, the control tower, and the warehouse operating system includes at least one or more of a computer-implemented training method, a computer-implemented inference method, a training system, an inference system, or combination thereof for training and inferring first and second neural networks, such as deep neural networks, to determine unplanned and unanticipated events that may cause a disruption within a supply chain ecosystem 100.

[0043] In one embodiment, an exemplary master intelligent network, configured to solve complex problems, may be located in a central, integrator site, and share its output as an input to one or more selected slave intelligent networks located at different remote locations. The master intelligent network will perform the training of neural network models and provide that training to the slave intelligent networks. In one embodiment, the master intelligent network is integrated into a supplier’s network, while individual ones of the slave intelligent networks are integrated into customer networks.

[0044] The master intelligent network is also able to receive and process the extrinsic data and then provide relevant portions of that processed extrinsic data to selected slave intelligent networks, which utilize that processed extrinsic data as an input and thereby increase their efficiency in performing the desired disruption predictions based on that processed extrinsic data (received from the master intelligent network). The slave intelligent networks would still be processing their own respective intrinsic data, which combined with the processed extrinsic data they’ve received, is used to predict anticipated disruptions in the supply chain. The slave intelligent network is also able to receive relevant, localized extrinsic data that the master intelligent network did not receive. This additional extrinsic data could be processed by the particular slave intelligent network and combined with, or utilized with, the processed extrinsic data (from the master intelligent network) in performing the prediction of possible or anticipated disruptions.

[0045] The centralized processing of extrinsic data at the master intelligent network, with the relevant portions of the processed extrinsic data shared with the respective slave intelligent works, eliminates the need for the individual slave intelligent networks toperform their own processing of extrinsic data. Such an efficiency can be seen in slave intelligent networks which are geographically close to each other and affected by the same local extrinsic events, such that their respective extrinsic datasets would be substantially the same, and thus prevent them from each having to separately process the same dataset of extrinsic data.

[0046] In another embodiment, an exemplary master intelligent network is configured as a quantum neural network that runs on quantum computing (e.g., computing with qubits, which can exist in multidimensional states) to solve complex problems. Furthermore, the neural networks could also be configured as quantum network networks.

[0047] If one of the slave intelligent networks is nearby another slave intelligent network, the slave intelligent network with a trained model can share its training with the other slave intelligent network and thereby eliminate the necessity of the other slave intelligent network needing to train its own model. In other words, the other slave intelligent network can use the trained model from the first slave intelligent network for prediction or anticipation of supply system disruptions (based upon a review of intrinsic and extrinsic data processed by the trained model(s)).

[0048] The intelligent network 160 can be implemented using software or a computer- implemented training and inference system and a method of executing in a cloud computing system or in any other appropriate manner for predicting a supply chain disruption within the supply chain ecosystem due to unplanned and unanticipated events over a time period such as near-range, short-range, or far-range, as depicted and described in further detail with respect to FIG. 3. The intelligent network 160 may further recommend actions or solutions and mitigate or reduce disruptions.

[0049] In another embodiment, the intelligent network 160 may be a cloud computing deep-learning system using cellular or radio access network, which may be Long Term Evolution (LTE), 5G, 6G, or any suitable cellular telecommunication protocol or radio access network.

[0050] In yet another embodiment, the intelligent network 170 includes an Internet and / or any appropriate local area networks (LANs), public switched telephone networks (PSTNs), packet data networks, optical networks, metropolitan area networks (MANs), wireless local area networks (WLANs), wide area networks (WANs), wired networks, wireless networks, any other suitable communication network, or any suitable combination of wired / wireless networks.

[0051] In addition to the above-described extrinsic data, there are intrinsic data induced by one or more of the facility entities 102, 104, 106, 108, 110, 112 of the intelligent supply chain ecosystem 100 that may also cause a disruption within the supply chain ecosystem 100, as depicted and described in further detail with respect to foregoing figures. For instance, the intrinsic data may be an unloading schedule, a reserve / storage schedule, a sortation / packing schedule, an order picking schedule, a loading status, robots schedule, vehicle schedule data, or other aggregated data for processing locally, externally, or remotely. Similar to extrinsic data, the intrinsic data either in real-time, stored data, or a combination thereof may be further trained and inferenced to determine unplanned and unanticipated events that may cause a disruption within the supply chain.

[0052] A computer-implemented training method, a computer-implemented inference method, a training system, an inference system, or combination thereof for training and inferring first and second neural networks, such as deep neural networks, for use as a deep training model and inference model to determine unplanned and unanticipated events that may cause a disruption within the supply chain ecosystem 100, as depicted and described in further detail with respect to FIG. 2.

[0053] In some cases, the computer-implemented training method, the computer- implemented inference method, the training system, the inference system, or combination thereof for training and inferring first and second neural networks, such as deep neural networks, for use as a deep training model and inference model can be implemented on a component such as a microprocessor, a computer processing unit, a graphical processing unit, a silicon-on-chip, a graphene-on-chip, a neural network-on-chip, a neuromorphic chip (NeuRRAM), a quantum computer performing quantum computing, a system-in-package (SIP) configuration, any suitable component, or any suitable combination of components, as depicted and described in further detail with respect to foregoing figures.

[0054] Referring to FIG. 2, a simplified schematic diagram of an intelligent supply chain ecosystem 200 according to another exemplary embodiment of the disclosure is illustrated. The intelligent supply chain ecosystem 200 includes a distribution network 204, numerous distribution centers 204a-204g in various geographical locations or in same geographical location are illustrated, the distribution center system 204 is communicatively coupled to an intelligent network 260 via communication links L4-1 , L4-2, L4-3, L4-4, L4-5, L4-5, L4-6, and L4-6, which may be wired, wireless, or any link suitable to support data communications and / or data packets. Any number of distribution centers may be formed as part of the supply chain ecosystem, according to particular needs.

[0055] The intelligent supply chain ecosystem 200 further includes a provider network 240 in various geographical locations different from other locations where the distribution network 204 is situated or in a same geographical location as the distribution network 204. The provider network 240 includes a weather center 240a, a transportation center 240b, a news station 240c, a social media center 240d, environment occurrences and natural disasters 240e, social unrest 240f, accidental event 240g, and any suitable service providers is communicatively coupled to the intelligent network 260 via communication links L6-1 , L6-2, L6-3, L6-4, L6-5, L6-6, L6-7 which may be wired, wireless, or any links suitable to support data communications and / or data packets. Inner and outer rings as depicted in FIG. 2 provides only an illustration and do not imply any limitations with regard to any specific location of the distribution network 204 and the provider network 240.

[0056] The communication links L4-1 , L4-2, L4-3, L4-4, L4-5, L4-5, L4-6, L4-6, L6-1 , L6-2, L6-3, L6-4, L6-5, L6-6, L6-7 may be wired, wireless, or any link suitable to supportdata communications and / or data packets between the distribution network 204 of the supply chain ecosystem 200, the provider network 240, and the intelligent network 260.

[0057] Extrinsic data produced by the provider network 240 may be in real-time. The data is then either stored as historic data or forward to the supply chain ecosystem 200 configured to train and inference data in real-time or at a later period. The data through training and inferencing may then be used to predict future events, recommend actions or solutions, mitigate disruptions, avoid disruptions, or the like. For example, the extrinsic data includes real-time weather data, historic weather data, forecast weather data, real-time traffic data, historic traffic data, forecast traffic data, real-time news feeds, historic news feeds, real-time social media feeds, historic social media feeds, real-time event data, historic event data, forecast event data, social unrest data, environment occurrences and natural disasters, or other events that is out of the control of the distribution network 204 of the intelligent supply chain ecosystem 200. The event as well as environment occurrences and natural disasters may be lockdown, pandemic, war, earthquakes, wildfires, cyclones, typhoons, tornadoes, floods.

[0058] In addition to the above-described extrinsic data, there are intrinsic data produced by one or more of the distribution centers 204a-204g located within the intelligent supply chain ecosystem 200. For instance, the intrinsic data may be an unloading schedule, a reserve / storage schedule, a sortation / packing schedule, an order picking schedule, a loading status, a robots schedule, a vehicle schedule data, or other aggregated data for processing locally, externally, or remotely. Similar to extrinsic data, the intrinsic data may also be produced in real-time, and either stored as historic data or forwarded to the intelligent network 160 and used to train and inference data in realtime, stored data, or combination thereof and predict future events. In some cases, the extrinsic and intrinsic data may be combined and used to train and inference data in real-time, stored data, or combination thereof and predict future events.

[0059] The intelligent network 260 comprises a data store or other data storage arrangement 262, a microprocessor 264, and a neural network (NN) 266. The data store 262 comprises an array of attribute data either retrieved from, sent by, transmittedby or any suitable methods, one or more of the distribution network 204 and the provider network 240. The attribute data may be extrinsic data, intrinsic data, or combination thereof. One data store 262 is illustrated in FIG. 2, it is understood that more than one data store may be incorporated into the intelligent network 260.

[0060] The microprocessor 264 can be referred as a system that performs data aggregation and effectuate process control functions, performs higher performance analytics, as well as training, and learning data collected by the provider network 240 without prior knowledge is operable to execute a computer program instruction stored on a memory (not shown). Although one microprocessor 264 is illustrated, one or more single-core and / or multi-core processors configured to perform artificial intelligent (Al) operations may be integrated into the intelligent network 260 without departing from the scope of the disclosure. The microprocessor 264 may be a digital signal processor (DSP), a general purpose core processor, a graphical processing unit (GPU), a computer processing unit (CPU), a microprocessor, an Al processing unit, an neural processing unit, a silicon-on-chip, a graphene-on-chip, a neural network-on-chip, a neuromorphic chip (NeuRRAM), a quantum computer performing quantum computing, a system on a chip (SoC), a system-in-package (SIP) configuration, or any suitable combination of components used for artificial neural network, quantum neural network, and / or reinforcement learning, as depicted and described in further detail with respect to foregoing figures.

[0061] Although in many cases, the data store 262, microprocessor 264, and NN 266 are formed as part of the intelligent network 260, in some embodiments, different network or hardware may be used to implement one or more of the data store 262, microprocessor 264, and NN 266.

[0062] In one embodiment, the intelligent network 260 includes an Internet and any appropriate local area networks (LANs), metropolitan area networks (MANs), wide area networks (WANs), any other suitable communication network, or any suitable combination of wired / wireless networks, configured to predict future events which may be near-range, short-range, or long-range disruptions. The intelligent network 260 mayfurther recommend actions or solutions to mitigate or avoid disruptions. The intelligent network 260, as depicted in FIG.2, is a cloud / on-premise enterprise intelligent network, other forms of cloud configuration are possible without deviate from the scope of the disclosure. Further yet, the intelligent network 260 is a NN that can be used to construct an input-output correspondence data model of a processing object in the disclosure.

[0063] The intelligent network 260 may be integrated into at least one or more of control towers, Warehouse Management System (WMS), Warehouse Execution System (WES), Warehouse Control System (WCS), or the like. In another embodiment, the intelligent network 260 is a standalone network and is communicatively coupled to at least one or more of control tower, Warehouse Management System (WMS), Warehouse Execution System (WES), or Warehouse Control System (WCS). In yet another embodiment, each supply chain ecosystem comprises a slave intelligent network and one or more of the slave intelligent networks may be communicatively coupled the provider network 240, the one or more supply ecosystems to a master intelligent network via communication links (not shown). In further embodiment, the intelligent network may be a mesh network, ad-hoc network, or suitable network suitable to support data communications and data packets.

[0064] The neural network (NN) 266 may be one or more of an Artificial Neural Network (ANN), Deep Neural Network (DNN), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Quantum Neural Network, Large Language Model (LLM), Multilayer Perceptron (MLP), Generative Adversarial Network (GAN), Fuzzy Neural Network (FNN), Spiking Neural Network (SNN), or the like. Alternatively, one or more of Reinforcement Learning (RL), Blockchain, and NN may be implemented in a combination manner. Further yet, more than two or more different type of NN may be implemented to the intelligent network 260.

[0065] The intelligent supply chain ecosystem 200 further includes a plurality of control towers 270a-270c having communication capability and coupled to the distribution network 204. In one embodiment, each control towers 270a-270c may be registeredand connected to a particular distribution center of the distribution network 204 within proximity. Data packets may be forwarded, transferred, and exchanged between control towers 270a-270c. In other embodiment, the control towers 270a-270c may use either mesh networking protocol, ad-hoc network protocol, star network protocol, or any available networking protocol suitable to support data communications and data packets. Similar to the intelligent network 260, the control towers 270a-270c may be constructed as a cloud-based system independently from the intelligent network. Alternatively, the control towers 270a-270c may be incorporated into the intelligent network 260 and form an intelligent control tower.

[0066] Referring to FIG. 3, the numeral 360 designates another embodiment of an intelligent network 360 for a supply chain ecosystem 300. Components of the intelligent network 360 may correspond with several components of intelligent network 260 described in reference to FIG. 2. For example, a provider network 340, a distribution network 304, and communication links L’1 -L'9 of FIG. 3 may correspond to provider network 240, distribution network 204, and communication links L4-1 to L4-7, L6-1 to L6-7 described in reference to FIG. 2. FIGs. 2 & 3 may illustrate similar intelligent supply chain ecosystems but are shown in different formats to provide a more detailed description of embodiments described herein.

[0067] The intelligent network 360 can be any infrastructure that allows both distribution network 304 and provider network 340 to share and access data and resources via communication links L’1 -L’10 which may be wired, wireless, or any link suitable to support data communications. Any number of distribution centers in various geographical locations may be formed as part of the supply chain ecosystem 300, according to particular needs. For example, the cloud-based deep learning intelligent network 360 may be used to predict future events, which may be near-range, short- range, or long-range disruptions, recommend actions or solutions, and / or mitigate or avoid disruptions. The cloud-based deep learning intelligent network 360 can be a public cloud accessible via an Internet by devices having Internet connectivity and appropriate authorizations to utilize. Alternatively, the cloud-based deep learningintelligent network 360 may be on a private cloud computing system, a hybrid cloud computing system, or any suitable computing system.

[0068] In another embodiment, the intelligent network 360 may comprise part of the cloud computing deep learning infrastructure or refer to a network existing in the cloud computing infrastructure. Here, the cloud-based deep learning intelligent network 360 may be constructed by using a cellular network, which may be 5G, 6G, Long Term Evolution (LTE), or any suitable cellular telecommunication protocol.

[0069] In yet another embodiment, the intelligent network 360 includes an Internet and any appropriate local area networks (LANs), metropolitan area networks (MANs), wide area networks (WANs), any other suitable communication network, or any suitable combination of wired / wireless networks.

[0070] The cloud-based deep learning intelligent network 360 comprises a database or other data storage arrangement 362, a processor 364, and a neural network (NN) 366. The database 362 comprises an array of attribute data either retrieved from, sent by, transmitted by or any suitable methods, one or more of the distribution networks 304 and the provider network 340. The attribute data may be extrinsic data, intrinsic data, or combination thereof. One database 362 is illustrated in FIG. 3, it is understood that more than one database may be incorporated into the intelligent network 360.

[0071] In one embodiment, the intelligent network 360 may be integrated into at least one or more of a control tower 370, a Warehouse Management System (WMS), a Warehouse Execution System (WES), a Warehouse Control System (WCS), or the like.

[0072] In another embodiment, the intelligent network 360 is a standalone network communicatively coupled to at least one or more of control tower 370, Warehouse Management System (WMS), Warehouse Execution System (WES), Warehouse Control System (WCS), or the like directly or indirectly.

[0073] In yet another embodiment, each supply chain ecosystem comprises a slave intelligent network and one or more of the slave intelligent networks may be communicatively coupled to the provider network 340, the one or more supplyecosystems coupled to a master intelligent network via communication links (not shown). As illustrated in FIG. 3, a control tower 370 also implemented using cloudbased communication is communicatively coupled to the cloud-based deep learning intelligent network 370.

[0074] The neural network (NN) 366 may be one or more of an Artificial Neural Network (ANN), Deep Neural Network (DNN), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Quantum Neural Network, Large Language Model (LLM), Multilayer Perceptron (MLP), Generative Adversarial Network (GAN), Fuzzy Neural Network (FNN), Spiking Neural Network (SNN), or the like. Alternatively, one or more of Reinforcement Learning (RL), Blockchain, and NN may be implemented in a combination manner. Further yet, more than two of different type of NN may be implemented to the intelligent network 360.

[0075] In yet another embodiment, each supply chain ecosystem comprises a slave intelligent network and one or more of the slave intelligent networks may be communicatively coupled to the provider network 340, the one or more supply ecosystems coupled to a master intelligent network via communication links (not shown). As illustrated in FIG. 3, a control tower 370 also implemented using cloudbased communication is communicatively coupled to the cloud-based deep learning intelligent network 370.

[0076] According to yet another embodiment, a master cloud-based deep learning intelligent network 460 for a supply chain ecosystem 400 may be used (FIG. 4). The master cloud-based deep learning intelligent network 460 is communicatively coupled to a plurality of slave cloud-based intelligent networks 460b-460n via communication links L which in turn are communicatively coupled to a plurality of distribution networks 404 (four distribution centers 404a-404n are illustrated) via communication links L. The communication links L may be wired, wireless, or any link suitable to support data communications. Any number of distribution centers or other suitable facility entities may be formed as part of the supply chain ecosystem 400, according to particular needs. For example, the cloud-based intelligent networks 460a-460n may be used topredict future events which may be near-range, short-range, or long-range disruptions, and recommend actions or solutions, and mitigate or avoid disruptions. The cloudbased intelligent networks 460a-460n can be a public cloud accessible via an Internet by devices having Internet connectivity and appropriate authorizations to utilize. Alternatively, the cloud-based intelligent networks 460a-460n may be on a private cloud computing system, a hybrid cloud computing system, or any suitable computing system.

[0077] In another embodiment, the intelligent networks 460a-460n may be on a cloud computing system using cellular or radio access network, which may be 5G, 6G, or any suitable cellular or radio access protocol.

[0078] In one embodiment, the intelligent network 460 may be integrated into at least one or more of control tower (not shown), Warehouse Management System (WMS), Warehouse Execution System (WES), Warehouse Control System (WCS), or the like.

[0079] In another embodiment, the intelligent network 460 is a standalone network communicatively coupled to at least one or more of control tower 470, Warehouse Management System (WMS), Warehouse Execution System (WES), Warehouse Control System (WCS), or the like directly or indirectly.

[0080] The master cloud-based deep-learning intelligent network 460 comprises a database 462, or other data storage arrangement 462a-n, a processor 464, and a neural network (NN) 466. The database 462 comprises an array of attribute data either retrieved from, sent by, transmitted by or any suitable methods, one or more of the distribution network 404 and the provider network (not shown). The attribute data may be extrinsic data, intrinsic data, or combination thereof.

[0081] The neural network (NN) 466 may be one or more of an Artificial Neural Network (ANN), Deep Neural Network (DNN), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Quantum Neural Network, Large Language Model (LLM), Multilayer Perceptron (MLP), Generative Adversarial Network (GAN), Fuzzy Neural Network (FNN), Spiking Neural Network (SNN), or the like. Alternatively, one or more of Reinforcement Learning (RL), Blockchain, and NN may be implemented in acombination manner. Further yet, more than two different types of NN may be implemented to the intelligent network 460.

[0082] In one embodiment, the master cloud-based deep-learning intelligent network 460 is located in a central, integrated site and shares its extrinsic data processing to one or more of the plurality of slave cloud-based intelligent networks 460b-460n (via communication links L), which are communicatively coupled to the plurality of distribution networks 404a-404n, which may be located in remote locations with respect to each other and the master cloud-based deep-learning intelligent network 460. The master cloud-based deep-learning intelligent network 460 will perform the training of neural network models and provide that training to the plurality of slave cloud-based intelligent networks 460b-460n. In one embodiment, the master cloud-based deeplearning intelligent network 460 is integrated into a supplier’s network, while individual ones of the plurality of slave cloud-based intelligent networks 460b-460n are integrated into customer networks.

[0083] The master cloud-based deep-learning intelligent network 460 is configured to receive and process the extrinsic data and provide relevant portions of that processed extrinsic data to selected ones of the plurality of slave cloud-based intelligent networks 460b-460n, which utilize that processed extrinsic data as an input and thereby increase their efficiency in performing the desired disruption predictions (at the respective distribution networks 404a-404n) based on that processed extrinsic data (received from the master cloud-based deep-learning intelligent network 460). In one embodiment, individual ones of the plurality of slave cloud-based intelligent networks 460b-460n are configured to perform processing on their own respective intrinsic data, which when combined with the processed extrinsic data received from the master cloud-based deep-learning intelligent network 460, is used to predict anticipated disruptions in the supply chain (e.g., with respective to individual ones of the distribution networks 404a- 404n). In another embodiment, individual ones of the plurality of slave cloud-based intelligent networks 460a-460n are configured to receive relevant localized extrinsic data that the master cloud-based deep-learning intelligent network 460 did not receive. This additional extrinsic data could be processed by the particular slave cloud-basedintelligent networks 460a-460n, and combined with, or utilized with, the processed extrinsic data (from the master cloud-based deep-learning intelligent network 460) in performing the prediction of possible or anticipated disruptions at the relevant distribution networks 404a-404n.

[0084] The centralized processing of extrinsic data at the master cloud-based deeplearning intelligent network 460, with relevant portions of the processed extrinsic data shared with respective slave cloud-based intelligent networks 460a-460n, eliminates the need for the individual slave cloud-based intelligent networks 460a-460n to perform their own processing of extrinsic data. Such an efficiency can be seen in slave cloudbased intelligent networks 460a-460n which are geographically close to each other and therefore affected by the same local extrinsic events, such that their respective extrinsic datasets would be substantially the same, and thus prevent them from each having to separately process the same dataset of extrinsic data.

[0085] In another embodiment, an exemplary master cloud-based deep-learning intelligent network 460 is configured as a quantum neural network that runs on quantum computing (e.g., computing with qubits, which can exist in multidimensional states) to solve complex problems. Furthermore, the neural networks could also be configured as quantum network networks.

[0086] If one of the slave cloud-based intelligent networks 460a-460n is geographically near another of the slave cloud-based intelligent networks 460a-460n, the slave cloudbased intelligent network (e.g., slave cloud-based intelligent network 460a) with a trained neural network model can share its training with the other slave cloud-based intelligent network (e.g., slave cloud-based intelligent network 460b) and thereby eliminate the necessity of the other slave cloud-based intelligent network (e.g., slave cloud-based intelligent network 460b) needing to train its own model. In other words, the slave cloud-based intelligent network 460b can use the trained model from slave cloud-based intelligent network 460a for prediction or anticipation of supply system disruptions (based upon a review of the intrinsic and extrinsic data processed by the trained model(s)).

[0087] Referring to FIG. 5, the numeral 500 designates another of an intelligent supply chain ecosystem 500 according to various embodiments. The intelligent supply chain ecosystem 500 includes a distributor network 504, a provider network 540, and an intelligent network 560 interconnected to each other via communication links 586, 588. Distributor network 504 may include one or more distribution centers 504a- 504n that may perform one or more services, tasks, or operations including, but not limited to, warehouse control operation; robotic and / or vehicle control operation for controlling robots 534a-534n and / or vehicles 536a-536n in the distributor network 504 and / or autonomous / automated robots and / or vehicles; tasks for stock / order replenishment and / or storage of inventory within the distributor network 504 and / or any facility entities optionally networked with the intelligent supply chain ecosystem 500; operations for coordinating freight, transportation, and shipping transactions including last miles delivery; operations for coordination of packing, picking, loading, unloading, and transferring articles; operations for coordination of cleaning, upkeeping, repairing, and inspecting the distribution center network 504; operations for coordination of prioritizing, rescheduling, monitoring / tracking, reporting, updating, alerting, and transmitting work orders, changes, status; and any suitable services.

[0088] Each distribution center 504a- 504n includes a receiving area or unloading dock 522a-522n, a reserve / storage area 524a-524n, a sortation / packing station 526a-526n, an order picking station 528a-528n, a shipping / dispatch area or loading dock 530a- 530n, robots 534a-534n, vehicles 536a-536n, and edge computing 532a-532n, and arecommunicatively coupled to each other via communication links or buses 582, 584. Other suitable Industrial Internet of Things (IIOTs) or internet of Things (lOTs) may be remotely connected and / or temporary provided to the distribution network 504. For instance, wearable device (eyeglasses, goggles, rings), user equipment (UE) previously described and depicted in FIG. 1 , and so forth may be used. The robots 534a-534n may be autonomous mobile robots (AMRs), autonomous guided vehicles (AGVs), autonomous aerial vehicles (drones), cleaning tote robots, automated mobile robots, automated guided vehicles, or any suitable robots. The vehicles 536a-536n may be forklifts, trucks, or any suitable vehicles having a software module or modules to at perform various tasks or operations such as using a controller system, using self-learning capability having an artificial learning module such as a neural network 566 connected to the software module or as part of the software module, using selfcorrecting capability to update tasks or operations in order to accommodate recommendations forwarded by the intelligent network 560, as depicted and described in further detail below.

[0089] The provider network 540 may include, but not limited to, a weather station 540a, a traffic station 540b, a news station 540c, and social media 540d. Each provider 540a, 540b, 540c, and / or 540d may generate an array of datasets, extrinsic data, or external forces and transmit to the data stores 562a-562n via communication link 588. For example, the extrinsic data may include weather data, traffic data, data from news feeds, data from social media feeds, social unrest data, environment occurrence and natural disaster data (such as lockdown, pandemic, war, earthquakes, wildfires, cyclones, typhoons, tornadoes, floods), or other suitable data.

[0090] In some cases, a plurality of service providers may operate together as a single provider network 540 and can be located in the same region. In other cases, the supply chain ecosystem 500 may include interconnected multiple provider networks located in various or different geographical locations / areas. The extrinsic data may be induced or generated in real-time, either stored as historic data or forwarded as real-time / current data to the intelligent network 560. The extrinsic data is either in real-time, stored data, or a combination thereof that may be further trained and inferenced to determine unplanned and unanticipated events that may cause a disruption within the supply chain.

[0091] The intelligent network 560 includes, but not limited to, one or more databases or data stores 562a-562n, a processor 564, a neural network 566, and a control tower 570. Other computer device or virtual computing device such as memory for various programs and data for the operation of the intelligent network 560 may be part of the intelligent network 560 or connected wirelessly to the intelligent network 560. Here the intelligent network 560 is illustrated as a cloud-based deep learning network. The data stores 562a-562n may include at least one or more of an array of training datasets, newdata, or combination thereof generated by at least one of the provider network 540, distributed network 504, or any facility entities optionally networked with the supply chain ecosystem 500. The data stores 562a-562n may further hold other types of data. For example, the data stores 562a-562n may hold raw or unstructured data gathered by at least one of the provider network 540, the distribution network 504, or any facility entities optionally networked with the supply chain ecosystem. The intelligent network 560 may be configured as an intelligent module, an intelligent component, an intelligent layer, an artificial training model, an inference model, or any suitable intelligent network including a computer-implemented training method, a computer-implemented inference method, a training system, an inference system, or combination thereof for training and inferring first and second neural networks to determine unplanned and unanticipated events that may cause a disruption within the supply chain ecosystem 500.

[0092] As depicted in FIG. 5, one neural network 566 is provided in the intelligent network 560. Further, the intelligent network 560 can include predictive logic to perform training and inferring one or more deep neural networks for predicting a supply chain disruption within the supply chain ecosystem due to unplanned and unanticipated events.

[0093] In one embodiment, the intelligent network 560 can be implemented using software or computer-implemented training and inference system and method executing in a cloud computing system or in any other appropriate manner for predicting a supply chain disruption within the supply chain ecosystem due to unplanned and unanticipated events over a time period such as near-range, short-range, or far-range. Also, the intelligent network 560 may recommend actions or solutions and mitigate or reduce disruptions.

[0094] In another embodiment, the intelligent network 560 may be a cloud computing system using cellular or radio access network which may be Long Term Evolution (LTE), 5G, 6G, or any suitable cellular telecommunication protocol or radio access network.

[0095] In yet another embodiment, the intelligent network 560 includes an Internet and / or any appropriate local area networks (LANs), public switched telephone networks(PSTNs), packet data networks, optical networks, metropolitan area networks (MANs), wireless local area networks (WLANs), wide area networks (WANs), wired networks, wireless networks, any other suitable communication network, or any suitable combination of wired / wireless networks.

[0096] In some cases, the intelligent network 560 is one of a system on a chip (SoC) or a system-in-package (SIP) configuration encompasses the data stores 562a-562n, the processor 564, the neural network 566, the control tower 570, the memory (not shown), and any suitable components are fabricated and packaged onto one single chip carrier.

[0097] In other cases, the intelligent network 560 includes a cloud infrastructure system that provides various services and these services are made available to any facility entities networked with the supply chain ecosystem directly or indirectly upon request.

[0098] The neural network 566 is configured to recognize an array of datasets and / or new data transmitted by one or more providers 540a-540d of the provider network 540, which may be one or more of an Artificial Neural Network (ANN), Deep Neural Network (DNN), Convolutional Neural Network (CNN), Deep CNN (DCN), Recurrent Neural Network (RNN), Quantum Neural Network, Grated RN (GRN), Large Language Model (LLM), Multilayer Perceptron (MLP), Generative Adversarial Network (GAN), Fuzzy Neural Network (FNN), Spiking Neural Network (SNN), Long Short Term Memory network (LSTM), Deep Stacking Network, Markov Chain, Bayesian Network (BN), Dynamic BN (DBN), Linear Dynamical System (LDS), Switching LDS (SLDS), Optical NN (ONN), or any suitable NN used as a deep training model and inference model to determine unplanned and unanticipated events that cause a supply chain disruption. Alternatively, one or more of Reinforcement Learning (RL), Deep RL (DRL), Blockchain, and NN may be implemented in a combination manner to form an intelligent network. Further yet, more than two different types of NN may be implemented to form the intelligent network 570.

[0099] As illustrated, the neural network 566 may be a DNN and use artificial learning model to learn from, categorize, and make predictions about a large amount of data generated by at least one of the provider network 540 and the distribution network 504.In other words, an artificial learning model is an extension of artificial intelligence capable of several tasks and / or operations including, but not limited to, classifying large amount of data generated by the provider network 540 and / or the distribution network 504 into multiple classes or level of status, clustering these input data into multiple groups or multiple types of disruption, determining an unplanned and unanticipated events, predicting a disruption due to the unplanned and unanticipated events, identifying patterns, trends, reoccurrences in input data, identifying a distribution of input data in a multiple-dimensional space, or combination of these.

[0100] The control tower 570 may be a cloud-based hybrid warehouse system and may host several warehouse management, execution, and controlling functions typically performed by WMS, WES, WCS.

[0101] Although one processor 564 is illustrated, one or more single-core and / or multi-core processors 564 configured to perform artificial intelligent (Al) operation may be integrated into the intelligent network 560 without departing from the scope of the disclosure. The processor 564 may be configured in communication with a memory (not shown) having a set of instructions used to program the processor 564 to perform various processes, tasks, or operations. Further, the processor 564 may enable to retrieve or extract data stored in the database 562a-n for training and / or inference, as depicted and described in further detail with respect to foregoing figures. The processor 564 may include any suitable processors. For example, the processor 564 may be a digital signal processor (DSP), a general purpose core processor, a graphical processing unit (GPU), a computer processing unit (CPU), a microprocessor, an Al processing unit, an neural processing unit, a silicon-on-chip, a graphene-on-chip, a neural network-on-chip, a neuromorphic chip (NeuRRAM), a quantum computer performing quantum computing, a system on a chip (SoC), a system-in-package (SIP) configuration, or any suitable combination of components used for artificial neural network, quantum neural network, and / or reinforcement learning, as depicted and described in further detail with respect to foregoing figures.

[0102] The communication links or buses 582, 584, 586, 588 may be wired, wireless, or any links / buses suitable to support data communications and data packets between the distribution network 504, the provider network 540, and the intelligent network 560.

[0103] Now back to the DNN 566, once a given network has been structured for a task the DNN 566 is trained using an array of training dataset(s) and / or an array of new data 562a-562n. For example, the training cycle can be performed in one or more of the training techniques including, but not limited to, supervised learning, unsupervised learning, semi-supervised learning, hybrid learning, incremental learning, distributed learning, reinforcement learning and combination thereof for training the dataset 562a- 562n.

[0104] In some cases, the training data in its raw form can be such as external forces, human induced activities, or the like generated by at least one of the provider network 540, the distribution network 504, and facility entity of the supply chain ecosystem and transmitted to the data stores 562a-562n for training the artificial learning model or pre-processed into another form, which can then further be utilized for training the same artificial learning model or a different artificial learning model. For example, the raw data can be smoothed, truncated, aggregated, clustered, or otherwise manipulated into another form, which can then be used for training the same or different artificial learning model. As illustrated in FIG.5, a computer-implemented training system, a computer-implemented inference system, a computer-implemented training method, a computer-implemented inference method, and a non-transitory computer- readable medium for training a deep neural network 566 for used as the artificial learning or training model to perform a task or operations are disclosed. Examples of tasks or operations include predicting supply chain disruptions, rescheduling order changes, rerouting shipments from one facility entity to another facility entity within the network and / or outside the network, prioritizing tasks to be performed by the robots and / or vehicles, prioritizing storage spaces to accommodate last minute inbound / receiving shipment, or any suitable tasks or operations.

[0105] As previously described, if the artificial learning model used for training the DNN 566 of the intelligent network 560 is responsible for predictive disruption, the dataset stored in the data store 562a-562n is labeled and classified into several level or groups, such as no delay, reasonable delay within a predetermined threshold, and extreme delay beyond the predetermined threshold. For example, the artificial learning model is capable of analyzing the dataset and providing a result that includes a classification of the dataset into a particular class, group, or level that may cause change in a warehouse inventory management operation.

[0106] In one embodiment, a warehouse inventory management system for managing / controlling the warehouse inventory management operation can be incorporated into the control tower 570.

[0107] In another embodiment, the warehouse inventory management system may be communicatively coupled to the control tower 570 either remotely or locally.

[0108] The artificial learning model can be continuously enhanced and take into account the impact of the changes in the warehouse operations triggered by external impacts / forces. If an output or the result corresponds to the delay exceeding the predetermined threshold or expected reasonable timeline, the artificial learning model may infer a disruption to a particular shipment. A WMS (not shown) based on output from the artificial learning model 566 of the intelligent network 560 can responsively order the receiving area 522a-522n affected by the disruption or impact of the changes to open up, unreserve, prioritize space at the receiving area 522a-522n to accommodate delay in the particular shipment. The artificial learning model can also send a message in various formats, such as text, audio, video, or combination thereof alerting the distribution center 504a-504n affected by the disruption to take one or more corrective measures. For instance, the message is sent to at least one or more of the robots 534a-534n and vehicles 536a-536n to, such as, clear or unreserve the storage space 524a-524n on storage racks in order to accommodate the delay in shipment. If the robots and vehicles are not available or are not within the proximity as detected by the control tower or the WMS, the message is then sent to a worker having amobile / wearable device proximal to the receiving area 522a-522n or storage area 524a- 524n to clear or make room in the storage space. The WMS continues to monitor the receiving pipeline and clear the next shipment. The WMS reserves the space on the racks for the next shipment and sends message to the other processing outbound units, going into transportation, logistics and yard to prepare for the next priority wave of shipment. When the disruption is restored to normal, and delayed shipment is cleared, the WMS reassesses the priority dynamically.

[0109] Referring to FIG. 6, the numeral 600 generally designates another embodiment of an intelligent supply chain ecosystem. Components of the intelligent supply chain ecosystem 600 may correspond with several components of the intelligent supply chain ecosystem 500 in FIG. 5. For example, a distribution network 604, a provider network 640, and an intelligent network 660 of FIG. 6 may correspond to distribution network 504, provider network 540, and intelligent network 560 of FIG. 5. FIGs. 5 and 6 may illustrate similar intelligent network but shown in different variations and provide a more detailed description of embodiments described herein.

[0110] Unlike the intelligent network 560 of FIG. 5, the intelligent network 660 further includes a reinforcement learning (RL) 668 communicatively coupled to the deep neural network (DNN) 666. The DNN 666 through RL 668 can learn whether an output corresponds to a result, i.e. a prediction of a disruption, level or disruption, or the like, and accordingly to learning whether the prediction is a true and correct statement. For example, DNN 666 receives features (not shown) as inputs and makes a prediction of a disruption and / or requests for recommendations corresponding to the features, and based on the prediction, the intelligent network 660 can produce signal indicative of results as outputs and transmit to a distribution center 604a-604n affected by the disruption. In some embodiments, the trained artificial learning model includes a DNN 666 using a least one or more deep reinforcement learning (DRL) 668 and a reward provided to the DNN 666 subsequent to the prediction of a disruption and / or requests for recommendations following the prediction. In another embodiment, the intelligent network 660 includes an integrated Reinforced Neural Network.

[0111] As best seen in FIG. 7 , the numeral 766 generally designates a neural network (NN) that can be trained to handle very large and complicated 3D dataset(s) in at least one embodiment of the disclosure. In the illustrated embodiment, the NN model 766 is divided into an input layer 792, an output layer 796, and intermediate layers 794, often referred to as hidden layers. Although three hidden layers are illustrated, it should be understood that there is no limit to a number of intermediate layers that can be utilized for processing a complex dataset by using NN model such as a deep neural network (DNN) model. A plurality of network nodes is positioned in different layers and connected via a plurality of connections C, which are within the scope of the disclosure. It should be noted that values on the nodes can be determined using different algorithms, which are also within the scope of the disclosure. The nodes, also referred as neurons, units, or cells, are depicted as circles. The nodes in the layers 792, 794, 796 include a particular representation of the data and are capable of performing weighted sum, max pooling, bypass, and other suitable types of operations. Each node in the hidden layers 794 may contain a multiplier, an adder, a comparator, and any suitable number of accumulators. With a NN, each hidden layer in the region 794 determines a non-linear transformation of a previous layer. The nodes B1 , B2, B3 within an area 798 are depicted as squares. The connections C may be weights or edges, defining as parameters of the neural network model 766. As depicted, each of the plurality of nodes xi-xnin layer 792 is coupled to each of the plurality of nodes in the hidden layers 794 through a plurality of corresponding connections C, defined as weights. Similarly, each of the plurality of nodes in the hidden layers 794 is coupled to each of the plurality of nodes yi-ys in the output layer 796 through a plurality of corresponding connections C, defined as weights. These corresponding weights and bias are trained weights and trained bias obtained through a training process according to artificial learning models, deep-learning methods, or any suitable learning models and methods. In some cases, the weight training can be performed on-chip using a memory / data store-on-chip or off-chip using a software.

[0112] During communication processing, input data of data sets 562a-562n, 662a-662n of FIGS. 5-6 may be fed to the nodes xi-xnas inputs in the input layer 792 through a training and inferencing model. For example, the training and inferencingmodel may be a computer-implemented training method, a computer-implemented inference method, a training system, an inference system, or combination thereof for training and inferring first and second neural networks, such as deep neural networks, for use as a deep training model and inference model to determine unplanned and unanticipated events that may cause a disruption within the supply chain ecosystem.

[0113] In one embodiment, the input data fed to the nodes xi-xnas inputs may include extrinsic data obtained from the provider network. In another embodiment, the input data fed to the nodes xi-xnas inputs may include intrinsic data obtained from one or more of the facility entities of the supply chain ecosystem. In yet another embodiment, the input data fed to the nodes xi-xnas input may include both extrinsic data and intrinsic data. The results through transformation / computation, i.e. after trained and inferred within the hidden layers 794, are outputted to the nodes yi-ys in the output layer 796 for performing various tasks or operations including, but not limited to, predicting a disruption. The input data, computational data, weights, results as output data may be stored on the memory / data store-on-chip.

[0114] The NN model 766 may be one or more of an Artificial Neural Network (ANN), Reinforced Neural Network, Deep Neural Network (DNN), Convolutional Neural Network (CNN), Deep CNN (DCN), Recurrent Neural Network (RNN), Quantum Neural Network, grated RN (GRN), Large Language Model (LLM), Multilayer Perceptron (MLP), Generative Adversarial Network (GAN), Fuzzy Neural Network (FNN), Spiking Neural Network (SNN), Long Short Term Memory network (LSTM), Deep Stacking Network, Markov Chain, Bayesian Network (BN), Dynamic BN (DBN), Linear Dynamical System (LDS), Switching LDS (SLDS), Optical NN (ONN), or the like. Alternatively, one or more of Reinforcement Learning (RL), Deep RL (DRL), Blockchain, and NN may be implemented in a combination manner to form an intelligent network. Further yet, more than two of different type of NN may be implemented to form an intelligent network.

[0115] FIG. 8 illustrates a simplified diagram of a deep neural network (DNN) 860 communicatively coupled to a deep reinforcement learning (DRL) 870 according to various embodiments of the disclosure. As previously described, the intelligent network870 includes a reinforcement learning (RL) 870 communicatively coupled to the neural network (NN) 866. The NN 866 through RL 870 can learn whether an output corresponds to a result, i.e. prediction of a disruption, level or disruption, or the like, accordingly to learning is true and correct statement. For example, NN 666 receives features (not shown) as inputs and makes a prediction of a disruption and / or requests for recommendations corresponding to the features, and based on the prediction, the intelligent network 860 can produce signal indicative of results as output and transmit to a distribution center affected by the disruption. The NN 866 may use a least one or more deep reinforcement learning 870 and a reward 890c provided to the NN 866 subsequent to the prediction of a disruption and / or requests for recommendations following the prediction.

[0116] For instance, reinforcement learning (RL) is designed to solve Markov decision chain (MDC) problems. Suppose there is an environment 870 at an arbitrary state 890d. An agent in the environment 870 performs an action 872. The action transits the environment 870 to the next state and the agent receives a reward 890c based on the outcome. The goal of MDC is to choose an action that maximizes the cumulative, discounted reward at any state. A Deep Q Network (DQN) model resolves large scale problems by approximating the expected value of the discounted, cumulative reward (Q value) using a NN 866. The action 872 includes no change 874, modify 876, alert 878, or any suitable actions. Example of actions may include rescheduling order changes, rerouting shipments from one facility entity to another facility entity within the network and / or outside the network, prioritizing tasks to be performed by the robots and / or vehicles, and prioritizing storage spaces to accommodate last minute inbound / receiving shipment, SMS message, etc. As previously described, if an output or the result corresponds to the delay exceeding the predetermined threshold or expected reasonable timeline, the artificial learning model may infer a disruption to a particular shipment and trigger one or more of a modify action 876 and an alert action 878. For instance, the alert action 878 in the form of a message is sent to at least one or more of the robots, the AGVs, the workers to clear or unreserve the storage space 524a, 524n on storage racks in order to accommodate the delay in shipment. The intelligent network 860 continues to monitor the receivingpipeline and clear the next shipment. The intelligent network 860 reserves the space on the racks for the next shipment and sends message to the other processing outbound units, going into transportation, logistics and yard to prepare for the next priority wave of shipment. When the disruption is restored to normal, and delayed shipment is cleared, the intelligent network 860 reassesses the priority dynamically.

[0117] FIG. 9A is a simplified block diagram illustrating a weight pattern of extrinsic data 940a-940n attributes of a three-dimensional (3D) deep neural network (DNN) in accordance with some embodiments of the disclosure. These data 940a-940n are acquired from the provider network 940. The 3D-DNN comprises an input layer, an array 3D hidden layers, an output layer. The weight pattern formed within the 3D-DNN includes multiple rows, three rows 905a-905c, are illustrated. These corresponding weights and bias are trained weights and trained bias obtained through a training process according to artificial learning models, deep-learning methods, or any suitable learning models and methods. It is noted that the weight pattern of extrinsic data 940a- 940n is merely an example and is not intended to limit the disclosure. Accordingly, it is understood that additional rows may be provided in between layers of the DNN. The extrinsic data attributes include real-time dataset, historic dataset, and forecast dataset. Each extrinsic data attribute in the illustrated embodiment includes weather data 940a, traffic data 940b, news feed 940c, social media feed 940d, event data 940n, and any suitable data acquired from a provider network. For instance, the traffic data 940b presented in 3D may include normal or flow traffic, dense traffic, congested traffic, or malicious traffic. Similarly, the weather data 940a also presented in 3D may include normal weather, severe weather, or extreme weather.

[0118] FIG. 9B is a simplified block diagram illustrating a weight pattern of intrinsic data 904a-940n-n attributes of a three-dimensional (3D) deep neural network (DNN) in accordance with some embodiments of the disclosure. These data 904a- 940n-n are acquired from the distribution center. The 3D-DNN comprises an input layer, an array 3D hidden layers, an output layer. The weight pattern formed within the 3D- DNN includes multiple rows, three rows 905a-905c, are illustrated. These corresponding weights and bias are trained weights and trained bias obtained through atraining process according to artificial learning models, deep-learning methods, or any suitable learning models and methods. It is noted that the weight pattern of intrinsic data 904a-940n-n is merely an example and is not intended to limit the disclosure. Accordingly, it is understood that additional rows may be provided in between layers of the DNN. The intrinsic data attributes include real-time dataset, historic dataset, and forecast dataset. Each intrinsic data attribute in the illustrated embodiment includes unloading data 904a, order fulfillment data 904b, robot feed 904c, vehicle feed 904d, loading data 904n-1 , map data 904n-n, and any suitable data acquired from a provider network. Map data from an external source may be acquired as needed during the process for a particular analysis performed by the intelligent network 460. For example, the external source is a global positioning system (GPS), geographic information system (GIS), LiDAR, etc.

[0119] FIGs. 10A-10B are simplified block diagrams illustrating a dual three- dimensional deep neural network (3D-DNN) 1060, 1060’ in accordance with various embodiments of the disclosure. Each of the 3D-DNN 1060, 1060’ comprises an input layer 1092, 1092’, multi-hidden layers 1094, 1094’, and an output layer 1096, 1096’. The 3D-DNN 1060, 1060’ not only enables to extract essential features in massive dataset collected from a provider network (not shown) using multi-layer artificial learning model, but it can also acquire essential features without requiring prior knowledge.

[0120] As depicted in FIGs. 10A-10B, an array of dataset 1102 is collected from the provider network 140 of FIG.1 and transmitted to the input layer 1092 of the 3D- DNN 1060, 1060’. The array of dataset 1102 comprises at least two groups of datasets generated by two service providers within the provider network 104. For instance, the first and second set of datasets (dataset 1102) may include a combination of datasets such as weather information / data, traffic information / data, news information, social media information, event data, and any suitable factors for predicting supply chain disruption in various formats such as text, image, video, meta-data, GPS coordinates, page links, tweets, author details, data, GPS coordinates, comments, CSS, or suitable machine / human readable formats. Here, the weather data comprises a geophysical condition W1 , a meteorological condition W2, and a climate condition Wn. Eachcondition W1 , W2, . . . , Wn may include features such as normal weather, severe weather, and extreme weather either short lived or over extended period of time. The traffic data comprises normal or flow traffic, dense traffic, and congested traffic. News feeds may include fake news, legitimate news, breaking news.

[0121] Alternatively, additional data may be collected include, but are not limited to, weather or climate data such as humidity, temperature, wind, precipitation, nature disaster data, etc. Each type of the data may be processed according to a unique data structure for that type of data. For instance, data, such as text data, collected from news or social media feeds may be processed with Natural Language Processing (NLP) model, image data collected from weather or traffic or transportation may be process with Convolution Neural Network (CNN) model, Recurrent Neural Network (RNN) model, or any suitable model.

[0122] Data aggregation for a given time period or over a period of time may include building a network and collapsing such as summarizing or combining all the information either extrinsic data, intrinsic data, or combination thereof that relates to that given time or period.

[0123] Once the data is aggregated at the input layer 1092, the intelligent network 960 may generate a weighted parameters for that given time or period at a first block a’ and then fuse the weighted parameters at a second block n’ using feature fusion model in a DNN. The first and second blocks a’, n’ formed as a hidden layer 1094. In some cases, more than two blocks may be form within the hidden layers without departing from the scope of the disclosure. The featured fused parameters at second block n’ is then outputted to the output layer 1096 having x1 , x2, x3, x4, x5, x6 (1104). As illustrated in FIG. 10B, the data x1 -x6 (1104) generated at the output layer 1096 of the first NN 1060 are now combined with other data (1202) such as supply chain (SC) data, map data and feed into input layer 1092’ (input data comprising dataset 1104 and the supply chain data, map data 1202) for training using a second NN 1060’. Based on second trained data 1204 generated and outputted to output layer 1096’having y1 -y5 (1204). The output y1 -y5 (1204) may be one or more of unplanned and unanticipated events, tasks or actions such as recommendations, messages, etc.

[0124] As previously described, a computer system described with reference to the figures herein may generally comprise a processor, an input device coupled to the processor, an output device coupled to the processor, and memory devices each coupled to the processor. The processor may perform computations and control the functions of the system, including executing instructions included in computer code for the tools and programs capable of implementing methods for monitoring warehouses, distribution centers, and intralogistics, in accordance with some embodiments, wherein the instructions of the computer code may be executed by the processor via a memory device. The computer code may include software or program instructions that may implement one or more algorithms for implementing one or more of the foregoing methods. The processor executes the computer code.

[0125] The memory device may include input data. The input data includes any inputs required by the computer code. The output device displays output from the computer code. A memory device may be used as a computer usable storage medium (or program storage device) having a computer-readable program embodied therein and / or having other data stored therein, wherein the computer-readable program comprises the computer code. Generally, a computer program product (or, alternatively, an article of manufacture) of the computer system may comprise said computer usable storage medium (or said program storage device).

[0126] As will be appreciated by one skilled in the art, the disclosure may be a computer program product. Any of the components of the embodiments of the disclosure can be deployed, managed, serviced, etc. by a service provider that offers to deploy or integrate computing infrastructure with respect to embodiments of the inventive concepts. Thus, an embodiment of the disclosure discloses a process for supporting computer infrastructure, where the process includes providing at least one support service for at least one of integrating, hosting, maintaining and deploying computer-readable code (e.g., program code) in a computer system including one ormore processor(s), wherein the processor(s) carry out instructions contained in the computer code causing the computer system for generating a technique described with respect to embodiments. In another embodiment, the disclosure discloses a process for supporting computer infrastructure, where the process includes integrating computer- readable program code into a computer system including a processor.

[0127] Aspects of the disclosures are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0128] These computer-readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks. These computer-readable program instructions may also be stored in a computer-readable storage medium that can direct a computer, a programmable data processing apparatus, and / or other devices to function in a particular manner, such that the computer-readable storage medium having instructions stored therein comprises an article of manufacture including instructions which implement aspects of the function / act specified in the flowchart and / or block diagram block or blocks.

[0129] The computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer-implemented process, such that the instructions which execute on the computer, other programmable apparatus, or otherdevice implement the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0130] The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the blocks may occur out of the order noted in the Figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and / or flowchart illustration, and combinations of blocks in the block diagrams and / or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts or carry out combinations of special purpose hardware and computer instructions.

[0131] It will be appreciated that numerous various to the above-mentioned approaches are possible. Variations to the above approaches may, for example, include performing the above steps in a different order.

[0132] While certain exemplary embodiments have been described and shown in the accompanying drawings, it is to be understood that such embodiments are merely illustrative of are not restrictive on the broad disclosure, and that this disclosure is not limited to the specific constructions and arrangements shown and described, since various other modifications may occur to those of ordinary skill in the art. The description is thus to be regarded as illustrative instead of limited.

Claims

CLAIMSThe embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:1 . A system for a supply chain ecosystem comprising one or more computing devices having a processor and a memory storing computer-executable instructions that, when executed by the processor, cause the system to: obtain extrinsic training data for training; perform a first training of a first neural network model with the extrinsic training data to train the first neural network model to generate a first set of trained data; obtain intrinsic training data for training; perform a second training of a second neural network model with the first set of trained data and the intrinsic training data to train the second neural network model to generate a second set of trained data; obtain a second set of extrinsic data; during a first processing, process the second set of extrinsic data using the trained first neural network model to generate a first set of processed data; obtain a second set of intrinsic data; during a second processing, process the second set of intrinsic data and the first set of processed data using the trained second neural network to generate a second set of processed data; and based on the second set of processed data, determine one or more of unplanned and unanticipated events to a supply chain disruption.

2. The system of claim 1 , wherein the computer-executable instructions, when executed by the processor, cause the system to: transmit an alert message of the unplanned and unanticipated events; and generate one or more recommendations for the supply chain disruption.

3. The system of claim 2, wherein the one or more recommendations is selected from a group consisting of: warehouse control operation, machine control operation,robotic control operation, vehicle control operation, stock or order replenishment, storage space prioritizing operation, freight operation, transportation, maintenance operation, order rescheduling, and labor resource planning.

4. The system of claim 3, wherein the first and second neural network models are selected from a group consisting of: Artificial Neural Network (ANN), Deep Neural Network (DNN), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Quantum Neural Network, Large Language Model (LLM), Multilayer Perceptron (MLP), Generative Adversarial Network (GAN), Fuzzy Neural Network (FNN), Spiking Neural Network (SNN), and combination thereof.

5. The system of claim 4, wherein the system is a system on chip (SoC) having the processor and the memory.

6. The system of claim 5, wherein the one or more computing devices comprises an intelligent network, and wherein the intelligent network is communicatively coupled to at least one of the processor and the memory and is formed as a part of the SoC.

7. The system of claim 6, wherein the extrinsic data and the extrinsic training data are generated by a provider network and the intrinsic data and the intrinsic training data are generated by one or more of facility entities of the supply chain ecosystem.

8. The system of claim 7, wherein the computer-executable instructions, when executed by the processor, cause the system to: extract information from the extrinsic data and intrinsic data during the first processing and / or the second processing, wherein information extracted from the extrinsic data and intrinsic data includes one or more datasets selected from the group consisting of text, image, video, meta-data, page links, tweets, author details, data, GPS coordinates, comments, and CSS.

9. The system of claim 8, wherein the extrinsic data and the extrinsic training data are selected from a group consisting of: weather data, traffic data, news data, data from social media feeds, and event data.

10. The system of claim 9, wherein the intrinsic data and the intrinsic training data are selected from a group consisting of: unloading schedule, reserve / storage schedule, sortation / packing schedule, order picking schedule, loading status, robots schedule, and vehicle schedule.11 . The system of claim 10, wherein the extrinsic data and the intrinsic data are live data, and wherein the extrinsic training data and the intrinsic training data are one of live data, previously recorded data, and a combination of live data and previously recorded data.

12. A method for a supply chain ecosystem comprising: obtaining extrinsic training data for training; performing a first training of a first neural network model with the extrinsic training data to train the first neural network model to generate a first set of trained data; obtaining intrinsic training data for training; performing a second training of a second neural network model with the first set of trained data and the intrinsic training data to train the second neural network model to generate a second set of trained data; obtaining a second set of extrinsic data; during a first processing, processing the second set of extrinsic data using the trained first neural network model to generate a first set of processed data; obtaining a second set of intrinsic data; during a second processing, processing the second set of intrinsic data and the first set of processed data using the trained second neural network to generate a second set of processed data; and based on the second set of processed data, determining one or more of unplanned and unanticipated events to a supply chain disruption.

13. The method of claim 12 further comprising: transmitting an alert message of the unplanned and unanticipated events; and generating one or more recommendations for the supply chain disruption.

14. The method of claim 13, wherein the one or more recommendations is selected from a group consisting of: warehouse control operation, machine control operation, robotic control operation, vehicle control operation, stock or order replenishment, storage space prioritizing operation, freight operation, transportation, maintenance operation, order rescheduling, and labor resource planning.

15. The method of claim 14, wherein the first and second neural network model are selected from a group consisting of: Artificial Neural Network (ANN), Deep Neural Network (DNN), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Quantum Neural Network, Large Language Model (LLM), Multilayer Perceptron (MLP), Generative Adversarial Network (GAN), Fuzzy Neural Network (FNN), Spiking Neural Network (SNN), and combination thereof.

16. The method of claim 15, wherein generating, by one or more computing devices, one or more recommendations for the supply chain disruption.

17. The method of claim 16, wherein the one or more computing devices comprising: a processor; a memory storing computer-executable instructions to be executed by the processor; an intelligent network comprising first and second neural network model; and a control tower.

18. The method of claim 17, wherein the intelligent network is formed as part of the control tower.

19. The method of claim 18, wherein the extrinsic data and the extrinsic training data are generated by a provider network and the intrinsic data and the intrinsic training data are generated by one or more of facility entities of the supply chain ecosystem.

20. The method of claim 19 further comprising extracting information from the extrinsic data and intrinsic data during the first processing and / or the second processing, wherein information extracted from the extrinsic data and intrinsic data may include: text, image, video, meta-data, page links, tweets, author details, data, GPS coordinates, comments, and CSS.21 . The method of claim 20, wherein the extrinsic data and the extrinsic training data are selected from a group consisting of: weather data, traffic data, news data, data from social media feeds, and event data.

22. The method of claim 21 , wherein the intrinsic data and the intrinsic training data are selected from a group consisting of: unloading schedule, reserve / storage schedule, sortation / packing schedule, order picking schedule, loading status, robots schedule, vehicle schedule.

23. The method of claim 12, wherein the extrinsic data and the intrinsic data are live data, and wherein the extrinsic training data and the intrinsic training data are one of live data, previously recorded data, and a combination of live data and previously recorded data.

24. A non-transitory computer-readable medium having stored thereon executable instructions that, in response to execution, cause one or more computing devices to perform operations, the operations comprising: receiving, by the one or more computing devices, a first array of aggregated attribute training data; performing, by the one or more computing devices, a first training of an artificial learning model with the first array of aggregated attribute training data to train theartificial learning model to identify training data that includes information corresponding to a type of condition; receiving, by the one or more computing device, a first array of aggregated attribute data; identifying, by the one or more computing devices, second data from the first array of aggregated attribute data that corresponds to a type of condition, wherein the second data includes information that the artificial learning model was taught to identify in the first training; receiving, by the one or more computing devices, a second array of aggregated training attribute data associated with the identified training data; performing, by the one or more computing devices, a second training of an artificial learning model with both the identified training data and the second array of aggregated attribute training data to identify enhanced training data; receiving, by the one or more computing devices, a second array of aggregated attribute data associated with the second data; identifying, by the one or more computing devices, second enhanced data from the second array of aggregated attribute data associated with the second data, wherein the second enhanced data comprises includes information that the artificial learning model was taught to identify in the second training; classifying, by the one or more computing devices, a grade for the second enhanced data; determining, by the one or more computing devices, whether the grade for the second enhanced data is either below, within, or above a predetermined threshold to a disruption; and transmitting, by the one or more computing devices, an alert message corresponding to the disruption if the data is graded above the predetermined threshold.

25. The non-transitory computer-readable medium of claim 24, wherein the first arrays of aggregated attribute training data and aggregated attribute data are generated by a provider network, wherein the provider network is selected from a group consisting of: a weather station, a traffic station, a news station, and social media.

26. The non-transitory computer-readable medium of claim 24, wherein the second arrays of aggregated attribute training data and aggregated attribute data are generated by a supply chain ecosystem having one or more facility entities, wherein the facility entity is selected from a group consisting of: a supplier, a manufacturer, a distribution center, a warehouse, and an e-commerce.

27. The non-transitory computer-readable medium of claim 24, wherein the artificial learning model is selected from a group consisting: Artificial Neural Network (ANN), Deep Neural Network (DNN), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Quantum Neural Network, Large Language Model (LLM), Multilayer Perceptron (MLP), Generative Adversarial Network (GAN), Fuzzy Neural Network (FNN), Spiking Neural Network (SNN), and combination thereof.

28. The non-transitory computer-readable medium of claim 24, wherein the alert message comprises one or more recommendations to the disruption.

29. The non-transitory computer-readable medium of claim 28, wherein the one or more recommendations is selected from a group consisting of: warehouse control operation, machine control operation, robotic control operation, vehicle control operation, stock or order replenishment, storage space prioritizing operation, freight operation, transportation, maintenance operation, order rescheduling, and labor resource planning.

30. The non-transitory computer-readable medium of claim 29, wherein the one or more computing devices comprising a processor, a memory, a data store, and a control towel.31 . The non-transitory computer-readable medium of claim 30, wherein the one or more computing devices is at least one of a cloud-based artificial computing device or a virtual computing device.

32. The non-transitory computer-readable medium of claim 31 , wherein the one or more computing devices are packaged into a single system on a chip (SoC).

33. A method for training an intelligent system for a supply chain ecosystem, the method comprising: processing first and second datasets with a first neural network of the intelligent system using a deep learning model; outputting a first trained output learned from first and second datasets; adding the first trained output into an input layer of a second neural network of the intelligent system; processing a third dataset and the first trained output with the second neural network using a feature learning model; and outputting a second trained output, wherein the second trained output includes (i) at least one unplanned and unanticipated event and (ii) a solution to the unplanned and unanticipated event.

34. The method of claim 33 further comprising determining data corresponding to the unplanned and unanticipated event, wherein the data includes a value either above, below, or within a predetermined threshold.

35. The method of claim 34 further comprising generating an alert message for the solution to the unplanned and unanticipated event if the value for the unplanned and unanticipated event is above the predetermined threshold.

36. The method of claim 35, wherein the first and second dataset are selected from a group consisting of: weather data, traffic data, news data, data from social media feeds, and event data.

37. The method of claim 36, wherein the third dataset is selected from a group consisting of: unloading schedule, reserve / storage schedule, sortation / packing schedule, order picking schedule, loading status, robots schedule, vehicle schedule.

38. The method of claim 37, wherein: the first and second datasets correspond to first time interval at first location; and the first trained output and the third dataset correspond to second time interval at second location.

39. The method of claim 38, wherein the first and second neural networks are selected from a group consisting of: Artificial Neural Network (ANN), Deep Neural Network (DNN), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Quantum Neural Network, Large Language Model (LLM), Multilayer Perceptron (MLP), Generative Adversarial Network (GAN), Fuzzy Neural Network (FNN), and Spiking Neural Network (SNN).

40. The method of claim 33, wherein the first, second, and third datasets are one of live data, previously recorded data, and a combination of live data and previously recorded data.41 . A system-on-package for supply chain disruption and configured to support a plurality of data types, the system-on-package comprising: a processor system comprising first and second processors; a non-transitory computer-readable medium on which is stored instructions for execution by the processor system, the stored instructions configured to: process first and second datasets with the first processor using a deep learning model trained with training datasets; output a first set of processed data from first and second datasets; add the first output into the second processor; process a third dataset and the first set of processed data with the second processor using a feature learning model trained with training datasets; and output a second set of processed data, wherein the second set of processed data includes (i) at least one unplanned and unanticipated event and (ii) a solution to the unplanned and unanticipated event.

42. The system-on-package of claim 41 , wherein the system-on-package is further configured to determine data corresponding to the unplanned and unanticipated event, wherein the data includes a value either above, below, or within a predetermined threshold.

43. The system-on-package of claim 42, wherein the system-on-package is further configured to generate an alert message for the solution to the unplanned and unanticipated event if the value for the unplanned and unanticipated event is above the predetermined threshold.

44. The system-on-package of claim 43, wherein the first and second datasets are selected from a group consisting of: weather data, traffic data, news data, data from social media feeds, and event data.

45. The system-on-package of claim 44, wherein the third dataset is selected from a group consisting of: unloading schedule, reserve / storage schedule, sortation / packing schedule, order picking schedule, loading status, robots schedule, vehicle schedule.

46. The system-on-package of claim 45 wherein: the first and second dataset correspond to a first time interval at a first location; and the first set of processed data and the third dataset correspond to a second time interval at a second location.

47. The system-on-package of claim 46, wherein the first and second neural networks are selected from a group consisting of: Artificial Neural Network (ANN), Deep Neural Network (DNN), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Quantum Neural Network, Large Language Model (LLM), Multilayer Perceptron (MLP), Generative Adversarial Network (GAN), Fuzzy Neural Network (FNN), and Spiking Neural Network (SNN).

48. A distributed intelligent system for a supply chain ecosystem comprising a plurality of distribution networks, the system comprising: a master intelligent network; a plurality of slave intelligent networks each communicatively coupled to at least one respective distribution network of the plurality of distribution networks; a provider network communicatively coupled to the master intelligent network and operable to provide a first set of extrinsic data to the master intelligent network; wherein the master intelligent network comprises a computer with a master neural network model operable to process the first set of extrinsic data to output a first set of processed data, wherein the master intelligent network is operable to select respective subsets of processed data from the first set of processed data to provide to selected ones of the plurality of slave intelligent networks; wherein a first slave intelligent network of the plurality of slave intelligent networks comprises a computer and a second slave neural network model operable to receive and process a respective first subset of processed data and a first set of intrinsic data related to a corresponding first distribution network to provide a first set of enhanced data; and wherein the first set of enhanced data includes at least one unplanned and unanticipated event affecting the first distribution network and a solution to the unplanned and unanticipated event.

49. The system of claim 48, wherein the provider network is communicatively coupled to each of the plurality of slave intelligent networks, and wherein the provider network is operable to provide a second set of extrinsic data to selected respective ones of the plurality of slave intelligent networks including the first slave intelligent network.

50. The system of claim 49, wherein the first slave intelligent network comprises a first slave neural network model operable to process a received second set of extrinsic data to output a second set of processed data, wherein the second set of processed data is different from the first subset of processed data.51 . The system of claim 50, wherein the second slave neural network model of the first slave intelligent network is operable to process the second set of processed data while processing the first subset of processed data and the first set of intrinsic data to provide the first set of enhanced data.

52. The system of claim 48, wherein the extrinsic data is selected from a group consisting of: weather data, traffic data, news data, data from social media feeds, and event data.

53. The system of claim 48, wherein the intrinsic data is selected from a group consisting of: unloading schedule, reserve / storage schedule, sortation / packing schedule, order picking schedule, loading status, robots schedule, vehicle schedule.

54. The system of claim 48, wherein the computer of the master intelligent network is configured to train the master neural network model to process the extrinsic data, such that the processed data comprises a relevant portion of the extrinsic data corresponding to an unplanned and unanticipated event.

55. The system of claim 48, wherein the computer of the master intelligent network is configured to train a first slave neural network model and a second slave neural network model and provide them to a selected slave intelligent network of the plurality of slave intelligent networks.

56. The system of claim 48, wherein the computer of the first slave intelligent network is configured to forward a trained first slave neural network model and a trained second slave neural network model to a second slave intelligent network of the plurality of slave intelligent networks.

Images