GB2622004A - Method and system for managing industrial processes - Google Patents
Method and system for managing industrial processes Download PDFInfo
- Publication number
- GB2622004A GB2622004A GB2212522.3A GB202212522A GB2622004A GB 2622004 A GB2622004 A GB 2622004A GB 202212522 A GB202212522 A GB 202212522A GB 2622004 A GB2622004 A GB 2622004A
- Authority
- GB
- United Kingdom
- Prior art keywords
- industrial process
- data
- model
- input data
- change
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 210
- 238000000034 method Methods 0.000 title claims abstract description 80
- 230000008859 change Effects 0.000 claims abstract description 87
- 230000008569 process Effects 0.000 claims abstract description 41
- 230000009471 action Effects 0.000 claims abstract description 24
- 230000004044 response Effects 0.000 claims abstract description 8
- 230000000977 initiatory effect Effects 0.000 claims abstract description 3
- 238000004422 calculation algorithm Methods 0.000 claims description 27
- 238000012545 processing Methods 0.000 claims description 15
- 238000013479 data entry Methods 0.000 claims description 13
- 238000010801 machine learning Methods 0.000 claims description 10
- 239000000463 material Substances 0.000 claims description 9
- 230000003044 adaptive effect Effects 0.000 claims description 7
- 238000004806 packaging method and process Methods 0.000 claims description 5
- 238000007639 printing Methods 0.000 claims description 5
- 230000000903 blocking effect Effects 0.000 claims description 4
- 239000002699 waste material Substances 0.000 claims description 4
- 238000012549 training Methods 0.000 claims description 3
- 238000010438 heat treatment Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000007637 random forest analysis Methods 0.000 description 2
- 239000002994 raw material Substances 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 239000000243 solution Substances 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 241000700605 Viruses Species 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009172 bursting Effects 0.000 description 1
- 238000001311 chemical methods and process Methods 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 230000034994 death Effects 0.000 description 1
- 231100000517 death Toxicity 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000001627 detrimental effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 238000010297 mechanical methods and process Methods 0.000 description 1
- 230000005226 mechanical processes and functions Effects 0.000 description 1
- 230000003278 mimic effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012858 packaging process Methods 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 238000000053 physical method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0259—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the response to fault detection
- G05B23/0286—Modifications to the monitored process, e.g. stopping operation or adapting control
- G05B23/0294—Optimizing process, e.g. process efficiency, product quality
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/048—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators using a predictor
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B17/00—Systems involving the use of models or simulators of said systems
- G05B17/02—Systems involving the use of models or simulators of said systems electric
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
- G05B19/41885—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by modeling, simulation of the manufacturing system
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0243—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model
- G05B23/0254—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Manufacturing & Machinery (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Multi-Process Working Machines And Systems (AREA)
- General Factory Administration (AREA)
Abstract
A method for managing an industrial process, comprises: Obtaining information pertaining to the industrial process 102 and creating one or more digital twin(s) of the industrial process based on the information 104. Receiving first set of input data from a data source(s) 106, wherein the first set of input data is indicative of parameter(s) of the industrial process. Predicting a first set of output data by executing a predictive domain model based on the first set of input data and the digital twin(s) 108. Executing the predictive domain model(s) using the first set of output data and the first set of input data to generate a second set of input data, wherein the second set of input data is representative of a predicted state(s) of industrial process. Determining change(s) required in industrial process, based on the predicted state(s) and required state(s) of the industrial process. And initiating, in response to the second set of input data, a process action to implement a change for optimizing the industrial process.
Description
METHOD AND SYSTEM FOR MANAGING INDUSTRIAL PROCESSES
TECHNICAL FIELD
This invention relates to digital twins. In particular, though not exclusively, this invention relates to a method for managing an industrial process and a system for managing an industrial process.
BACKGROUND
With a boom in technological growth and population, various industries have sprouted over the world to meet growing demands of the consumers. Due to this surge in development of industries, there were initially discrepancies with respect to how industrial processes were carried out. In order to streamline such processes and standardise functioning of the industries, the industrial processes have been implemented. Each industry having its own specialisation generally has its own set of industrial processes which are utilised therein. Such industrial processes are often used for all industries including mechanical, electronic, chemical, manufacturing, packaging, oil and gas, and so forth.
However, often due to small developments within the industrial process, or wear-and-tear of physical machinery, or other such parameters, such industrial processes get derailed. In most cases, this leads to a wastage of raw materials, damage of machinery, and so forth. Such cases not only lead to wastage of economies which would otherwise be used elsewhere, but are also unsustainable for the environment. For example, oil pipes on a sea floor bursting not only increase costs of rebuilding the same, but also result in deaths of aquatic life.
Therefore, in light of the foregoing discussion, there exists a need to overcome the aforementioned drawbacks associated with managing industrial processes.
SUMMARY OF THE INVENTION
In a first aspect, an embodiment of the present disclosure provides a method for managing an industrial process, the method comprising: obtaining information pertaining to the industrial process; - creating at least one digital twin of the industrial process based at least on the information pertaining to the industrial process, wherein the at least one digital twin is dynamically updated while the industrial process is executed; - receiving a first set of input data from at least one data source, wherein the first set of input data is indicative of at least one parameter in the industrial process; - predicting a first set of output data by executing at least one predictive domain model that is pre-trained, wherein the first set of output data is predicted based on the first set of input data and the at least one digital twin, and wherein the first set of output data is representative of at least one predicted state of the industrial process; - executing the at least one predictive domain model that is pre-trained, using an addition of a first set of output data and the first set of input data, to generate a second set of input data, wherein the second set of input data is an accurate representation of the at least one predicted state of the industrial process; - determining at least one change required in the industrial process, based on the at least one predicted state and at least one required state of the industrial process; and - initiating, in response to the second set of input data, at least one process action to implement the at least one change, for optimizing the industrial process.
Throughout the present disclosure, the term "industrial process" refers to a process carried out in an industry. Such processes may be implemented for producing goods or providing services. Examples of the industrial process include, but are not limited to, a mechanical process, a chemical process, a mining process, an automation process, a manufacturing process, a printing process, a packaging process. Moreover, herein the term 'managing' pertains to all aspects of managing the industrial process, for example, observing the industrial process, detecting possible issues or drawbacks in the industrial process, finding solutions for the industrial process, optimizing the solutions for the industrial process, and so forth.
The term "information pertaining to the industrial process" refers to data indicative of various aspects of the industrial process. Notably, such information provides an insight into the industrial process. Optionally, such information is obtained using at least one sensor. Alternatively, optionally, such information is historical information from previously implemented industrial processes. In this case, the information is obtained from an external device, a database, and the like. Optionally, the information pertaining to the industrial process comprises at least one of: information pertaining to machinery involved for the industrial process, information pertaining to structure of the machinery, information pertaining to specifications of the machinery, information pertaining to raw materials, information pertaining to steps being followed for the industrial process.
Throughout the present disclosure, the term "digital twin" refers to a virtual representation of a real-world asset, which serves as a real-time digital counterpart. This means that, the at least one digital twin is updated in real-time during its entire lifecycle, depending on changes observed in the at least one real-world asset. Moreover, the at least one digital twin utilises at least simulation, machine learning and reasoning technologies for assisting in decision-making. Herein, the real-world asset refers to an object or a process which exists in a real-world. The at least one digital twin is created using the information pertaining to the industrial process. As previously mentioned, such information provides insights into the industrial process. It will be appreciated that the at least one digital twin is a virtual counterpart of the industrial process. Notably, the at least one digital twin is dynamically updated, which means that any changes being observed in the industrial process are updated in the at least one digital twin while the industrial process is being updated. The at least one digital twin being dynamically updated allows possible shortcomings in the industrial process to be identified before significant damage occurs, and thereby be resolved. Updating of the at least one digital twin comprises removing existing data values, and/or replacing the existing data values with new (i.e., updated) data values. It will be appreciated that such updating ensures that the at least one digital twin is a real-time virtual replica of the industrial process, since the at least one digital twin is constantly updated depending on changes of the industrial process. This allows functions to be performed on the at least one digital twin virtually, for improving performance and accuracy. Optionally, the at least one digital twin may not be updated in real-time. Herein, a process of the at least one digital twin may be slowed down to provide an output at desired time intervals, such that information may be derived from the at least one digital twin. In this manner, small changes may be captured which assist in optimising the industrial process. Moreover, the process of the at least one digital twin may be accelerated to anticipate behaviour and performance.
The term "input data" refers to a collection of data or information which is used as an input for the at least one predictive domain model. The first set of input data is used as an input for predicting the first set of output data using the at least one digital twin. The input data is optionally expressed as numerical values indicative of the at least one parameter at a given time instant. Optionally, the input data is presented in a tabular pattern, wherein each column corresponds to a given parameter and each row corresponds to a given time instant. Optionally, the input data is predictive of at least one state of the industrial process.
Optionally, the input data comprises at least one of: historic data, real-time data, of the industrial process. The historic data refers to collected data pertaining to past events that have transpired. For example, the historic data may comprise data pertaining to a drill for a mining industrial process. Optionally, the historic data is used for training the at least one predictive domain model for managing the industrial process. Beneficially, such historic data provides information regarding how the industrial process is conducted and issues that may have occurred in the past. Moreover, using such data allows the at least one predictive domain model to learn possible scenarios from historical occurrences. The real-time data refers to collected data pertaining to instantaneous changes in the industrial process. Such real-time data is instantaneously available as soon as it is created and/or acquired. Optionally, the real-time data is used for managing the industrial process. Beneficially, such real-time data provides information pertaining to the execution of the industrial process, such that if there are any discrepancies, the at least one predictive domain model can identify the same and suggest appropriate changes required.
The term "data source" refers to a memory which is configured to store at least the input data. Optionally, the at least one data source is implemented as at least one of: a virtual sensor, a physical sensor, a device. The virtual sensor is a sensor which is virtually deployed to sense the at least one parameter in the industrial process. Such virtual sensors are trained using historical data to sense the at least one parameter and are not placed physically within a system. Optionally, such virtual sensors are built using principles of physics and chemistry. It will be appreciated that the virtual sensor is often a result of a model. In some cases, deploying physical sensors (such as, for example, a flow sensor) is often avoided due to high costs involved. In such cases, virtual sensors are utilized, which assist in optimizing processes as well since they are built using the model. Herein, the model calculates a series of criteria (i.e., parameters), pertaining to a given process (such as, for example, temperature, pressure, flow, and so forth). In an example, the industrial process involves a set of interconnected floatation cells, wherein some parameters between individual floatation cells are not physically measured, and merely an intake of a first floatation cell and an output of a last floatation cell are measured. In such a case, these parameters are generated using virtual sensors for optimizing the industrial process.
The physical sensor is an electromechanical device which senses the at least one parameter in the industrial process. Such physical sensors are deployed by being physically placed within the industrial process. Examples of the physical sensor include, but are not limited to, a pressure sensor, a motion sensor, a light sensor, a flow sensor, a temperature sensor, an optical sensor, a magnetic sensor, a proximity sensor, an infrared sensor, a level sensor. The device is an electromechanical device capable of capturing, storing and/or sharing the input data. Herein, the device may be implemented as: a device of the industrial process, an external device, a user device. The device of the industrial process is a device embedded within the industrial process, the external device is a device which is not embedded within the industrial process but provides some insight into the industrial process, and the user device is a device associated with a user and which is capable of capturing, storing, and sharing data. Beneficially, the above-mentioned implementations of the at least one data source provide valuable insights into the industrial process.
Optionally, the at least one data source generates control signals for controlling the industrial process in real-time. The term "control signal' refers to a signal which represents a control command for controlling the industrial process. Optionally, the industrial process is controlled in near real-time. In such cases when real-time data cannot be instantaneously implemented using the at least one predictive domain model, the control signals for controlling the industrial process are generated by the at least one data source in near real-time.
Throughout the present disclosure, the term "parameter" refers to an element of the industrial process, which is useful for identifying the industrial process, or evaluating at least a performance, a status, or condition of the industrial process. Examples of the at least one parameter include, but are not limited to, a pressure parameter, a motion parameter, a light parameter, a flow parameter, a temperature parameter, an optical parameter, a magnetic parameter, a proximity parameter, an infrared parameter, a level parameter. Notably, such first set of input data is received such that corresponding first set of output data, and thereafter the second set of input data may be predicted using the at least one predictive domain model to determine the at least one change.
Optionally, prior to the step of predicting the first set of output data, the method comprises: - creating the at least one predictive domain model using at least one machine learning algorithm; and - training the at least one predictive domain model using the input data.
Throughout the present disclosure, the term "predictive domain model" refers to a model which predicts future events or outcomes in the industrial process by analysing patterns using the at least one digital twin of the industrial process and the input data. Optionally, a predictive a domain model can be implemented for simulating a portion or the whole of an industrial process. Examples of the at least one predictive domain model include, but are not limited to, a classification model, a clustering model, a forecast model, an outliers model, a time series model. Notably, the at least one predictive domain model is created using the at least one machine learning algorithm. The term "machine learning algorithm" refers to an algorithm which creates the at least one predictive domain model, such that it is able to learn and predict outcomes without being explicitly trained to do so. Moreover, the at least one predictive domain model is thereby trained using the input data, which is labelled, allowing the at least one predictive domain model to learn and grow accurate over time. Examples of the at least one machine learning algorithm include, but are not limited to, a linear regression algorithm, a logistic regression algorithm, a decision tree algorithm, a support vector machine (SVM) algorithm, a Naive Bayes algorithm, a k-nearest neighbors (KNN) algorithm, a K-means algorithm, a random forest algorithm. Optionally, the predictive domain model is implemented using artificial intelligence. Optionally, the predictive domain model is based on a predictive algorithm. Examples of the predictive algorithm include, but are not limited to, a random forest algorithm, a generalized linear model (GLM) algorithm, a gradient boosted model (GBM) algorithm, a K-means algorithm, a prophet algorithm. Beneficially, the at least one predictive domain model is created and trained herein such that the at least one predictive domain model is pre-trained for use in optimizing the industrial process. It will be appreciated that the at least one predictive domain model leverages the input data to optimize the industrial process.
The term "output data" refers to data which is outputted from the at least one predictive domain model. It will be appreciated that the first set of output data is outputted from the at least one predictive domain model, and the first set of input data is inputted into the at least one predictive domain model. Herein, the first set of output data is representative of the at least one predicted state of the industrial process based on the first set of input data. It may be appreciated that an output is an outcome of an executed process and similarly the predicted output data obtained herein is the outcome of execution of the at least one predictive domain model. Moreover, the at least one predictive domain model is a pre-trained model. Furthermore, the at least one predictive domain model may include, but not be limited to, a standard industrial process model, a tested industrial operation model, and so forth. In an example, the at least one predictive domain model may include a machine learning model and/or an artificial intelligence model to pre-train the at least one predictive domain model, which predicts the first set of output data and the second set of input data based on the execution of the at least one predictive domain model. Furthermore, the prediction of the first set of output data is based on the first set of input data and the at least one digital twin. It may be appreciated that the first set of input data provides a trigger and/or a throughput that simulates processing of the first set of output data. Herein the prediction of the first set of output data is a cumulative process of translating the first set of input data to the at least one digital twin and executing the at least one predictive domain model in tandem.
Throughout the present disclosure, the term "predicted state" refers to a state of the industrial process which is predicted based on the first set of input data. Furthermore, the at least one predicted state may be interpreted as a step of a conventional industrial process. Herein, the step of the conventional industrial process includes a portion of the step, a prior step which has already been executed, a future step which has to be executed. In an example, the at least one predicted state may include an increase in temperature of a heating process, a reduction in pressure in a compression process, and such like in the conventional industrial process. According to an embodiment of the present disclosure, the at least one predicted state is represented through the second set of input data, and the first set of output data of the industrial process. Based on an implementation with the present embodiment, the second set of input data is beneficially predicted prior to physical occurrence in the industrial process. Furthermore, the prediction is based on execution of the at least one predictive domain model and its representation using the second set of input data.
Optionally, the at least one predictive domain model comprises at least one of: a material model, a ball charge model, a liner wear model, a dynamic charge model, inferential trajectory model, ball breakage model, new ball ejection model, a material processing model, a data processing model, a waste management model, a packaging model, an assembly model, a printing model. It will be appreciated that the at least one predictive domain model auto generates determining steps based at least one of: a given input data, a given data source, an ideal industrial method, a reoccurrence, a combination thereof. Furthermore, the at least one predictive domain model when executed, defines the predictive state of the industrial process.
Optionally, the addition of the first set of output data and the first set of input data is implemented as: a mathematical addition of respective values, a mean of respective values. For example, if the first set of output data has a value 4 and the first set of input data has a value 5, then the addition would either be 9 or 4.5, depending on if the addition is implemented as a mere addition or a mean. It will be appreciated that the second set of input data is outputted from the at least one predictive domain model, and the addition of the first set of input data and the first set of output data is inputted into the at least one predictive domain model. Herein, the second set of input data is the accurate representation of the at least one predicted state, which means that the second set of input data entails parameters in accordance with parameters during the at least one predicted state. It will be appreciated that the second set of input data is utilised to determine the at least one change with respect to the at least one predicted state.
Optionally, the above-mentioned process is repeated until an acceptable value of a confidence factor is achieved. The confidence factor is indicative of an accuracy of prediction and/or function. The acceptable value of the confidence factor demonstrates efficient functioning of the industrial process. In this regard, it will be appreciated, that the at least one domain model is continually executed with updated values of the first set of input data until a required accuracy is achieved. Throughout the present disclosure, the term "at least one change" refers to an iteration in the industrial process, which results in a deviation of the overall industrial process. The term "required state" refers to a desired output considering the given change in the first set of input data that deviates an ongoing process to achieve the at least one required state. The at least one required state may indicate at least one of: an ideal state, a standard procedural state, a position, in the industrial process. Furthermore, the at least one required state may be referenced from a conventional industrial plan, or an established industrial data obtained from experimental setups and experimental data. Herein, the at least one predicted state and the at least one required state of the industrial process are mapped to find similarities and/or dissimilarities. Herein, the at least one change would be required to change the at least one predicted state to the at least one required state. This means that once the at least one change is implemented, the at least one predicted state and the at least one required state must be similar. When the at least one predicted state is similar to the at least one required state, the at least one predicted state will mimic the at least one required state, and provide outputs akin to a local sensor installed within the industrial process. For example, if the at least one predicted state is a temperature of 30 degrees and the at least one required state is a temperature of 100 degrees in a heating chamber, the at least one change may be to increase the heat within the heating chamber to reach 100 degrees.
Optionally, when determining the at least one change required in the industrial process, the method is configured to: - determining at least one proposed change based on the at least one predicted state and the at least one required state; - simulating the at least one proposed change in the at least one digital twin of the industrial process to generate a simulated output state; - assessing whether the simulated output state is similar to the at least one required state; and - when the simulated output state is similar to the at least one required state, implementing the at least one proposed change as the at least one change required in the industrial process.
The at least one proposed change refers to at least one change proposed by the at least one prediction domain model for changing the at least one predicted state to the at least one required state. The at least one proposed change is determined by mapping similarities and differences of the at least one predicted state and the at least one required state. In an example, the industrial process requiring a change in temperature by an increase of heat input in a heating chamber, amounts as at least one proposed change for the heating chamber of the industrial process. Thereon, the at least one proposed change is simulated in the at least one digital twin to anticipate an output state of the industrial process when the at least one proposed change is implemented. Since the at least one proposed change is simulated, it beneficially saves costs and effort while providing insights with respect to the at least one proposed change being simulated. The simulated output state is mapped with the at least one required state by comparing similarities and differences of the same. Notably, the at least one proposed change is implemented as the at least one change required in the industrial process only when the simulated output is similar to the at least one required state. It will be appreciated that such determination of the at least one change required in the industrial process is efficient, sustainable and saves excessive costs since it simulates the at least one proposed change in the at least one digital twin before applying the at least one change to the industrial process. Herein, if any issues are flagged during the simulation, those are sought out and the at least one proposed change is altered until it provides a desired result.
Since the second set of input data is predictive of at least one state of the industrial process, it is utilized to determine the at least one predicted state. It will be appreciated that the at least one state is an actual state of the industrial process in real-time, whereas the at least one predicted state is based on the prediction of the second set of input data. The term "process action" refers to an action which is to be performed in real-time in the industrial process. In some cases where the industrial process is physically deployed, the at least one process action is a physical action. In other cases where the industrial process is virtually deployed, the at least one process action is a virtual action. Examples of the at least one process action include, but are not limited to, changing a temperature, changing a pressure, changing a material supply rate, of the industrial process. Beneficially, such method optimizes the industrial method by efficiently determining the at least one change required in the industrial process and implementing the same using the at least one process action.
Optionally, the at least one process action is initiated autonomously or semi-autonomously. Herein, when the at least one process action is initiated autonomously, no approvals or verifications are required from any entity, however, when the at least one process action is initiated semi-autonomously, an approval and/or verification is required from an entity.
Optionally, the method is configured to employ adaptive data encryption and data obfuscation processing operations depending upon the at least one parameter of the industrial process. Herein, such adaptive data encryption and data obfuscation processing operations garner a degree of data protection to the method, making it safer against hacking attacks. Optionally, the data is being encrypted using any suitable method of data encryption. Optionally, data is randomized before encryption for increased security. By enhancing the degree of data protection, the method is less prone to being disrupted by malicious third parties, for example by injection of computer viruses or by selective eavesdropping and substitution of data flows. By employing data protection that is adaptively adjusted depending upon at least one parameter of the industrial process, the method can be both highly efficient in its operation and also highly robust to attack. Furthermore, the adaptive data encryption and data obfuscation advantageously make the method robust, and prevent it from unwanted intrusions, for example third-part malicious attacks. Moreover, adaptive data encryption employed by the given digital twin encrypts the data being exchanged based on the type of data. Moreover, such data includes a confidence factor of the given digital twin, the information pertaining to the industrial process, the first set of input data, the first set of output data, the second set of input data, the at least one change, and so forth.
Optionally, the method further comprises sending a notification to at least one device associated with an entity, wherein the notification is indicative of the at least one change, and wherein the entity performs the at least one change to optimize the industrial process. Herein, the at least one change is being implemented semi-autonomously since the at least one predictive domain model is being utilized to determine the at least one change but the at least one change is eventually performed by the entity. The term "entity" refers to a physical entity capable of performing the at least one change. Optionally, the entity is implemented as at least one of: a person, a robot. The at least one device associated with the entity refers to a communication device capable of receiving and accessing the notification. Optionally, the notification is implemented as at least one of: a visual notification, an audio notification, a haptic notification. It will be appreciated that the notification is sent to the at least one device associated with the entity, such that the entity performs the at least one change. If such a notification is not provided in a timely manner, it may be detrimental to the industrial process. Thereby, the sending of such notification to the entity is beneficial since it allows timely resolution of issues in the industrial process and avoids unnecessary wastage or damage.
Optionally, the method further comprises: - calculating an uncertainty of the at least one process action for implementing the at least one change; - comparing the uncertainty with a predefined uncertainty threshold; and - when the uncertainty is greater than the predefined uncertainty threshold, blocking at least one data entry from the first set of input data.
The term "uncertainty" refers to an epistemic situation involving imperfect or unknown information. Such uncertainty is applicable to predictions of future events, predetermined physical measurements, an unknown, and so forth. Optionally, the uncertainty is implemented as at least one of: a state uncertainty, an effect uncertainty, a response uncertainty. In operation, the uncertainty is caused by the at least one data entry of the first set of input data. The at least one data entry refers to a data entry of the first set of input data which causes a discrepancy and increases the uncertainty. Such data entry is often unreliable and thereby jeopardizes the prediction of the at least one change. Optionally, the at least one data entry is utilized to train the at least one predictive domain model. The uncertainty is calculated by mapping if the at least one process action would implement the at least one change. When excessive disparity is observed between the two, it may be assumed that the at least one change is unreliable and thereby a corresponding process action is not implemented. Beneficially, blocking the at least one data entry from the first set of input data removes the uncertainty since the uncertainty was being caused by the at least one data entry. Moreover, calculating the uncertainty and thereon blocking the at least one data entry is beneficial since it safeguards the industrial method against unreliable data.
Optionally, the method further comprises employing a monte carlo dropout algorithm to estimate the uncertainty. The monte carlo dropout algorithm refers to a class of computational algorithms that rely on repeated random sampling to obtain a distribution of a numerical quantity. Optionally, when the uncertainty is greater than the predefined uncertainty threshold, the method further comprises sending an alert to the at least one device associated with the entity. In such cases, the determination of the at least one change is halted, and the entity is to overlook the industrial process until the uncertainty is not reduced to a value equal to or lower than the predefined uncertainty threshold.
In a second aspect, an embodiment of the present disclosure provides a system for managing an industrial process, the system comprising at least one data source and at least one processor, wherein the at least one data source and the at least one processor are communicably coupled, and wherein the at least one processor is configured to: - obtain information pertaining to the industrial process; - create at least one digital twin of the industrial process based at least on the information pertaining to the industrial process, wherein the at least one digital twin is dynamically updated while the industrial process is executed; - receive a first set of input data from the at least one data source, wherein the first set of input data is indicative of at least one parameter in the industrial process; - predict a first set of output data by executing at least one predictive domain model that is pre-trained, wherein the first set of output data is predicted based on the first set of input data and the at least one digital twin, and wherein the first set of output data is representative of at least one predicted state of the industrial process; - executing the at least one predictive domain model that is pre-trained, using an addition of a first set of output data and the first set of input data, to generate a second set of input data, wherein the second set of input data is an accurate representation of the at least one predicted state of the industrial process; - determine at least one change required in the industrial process, based on the at least one predicted state and at least one required state of the industrial process; and - initiate, in response to the second set of input data, at least one process action to implement the at least one change, for optimizing the industrial process.
Throughout the present disclosure, the term "processor" refers to hardware, software, firmware, or a combination of these configured to control operation of the system. In this regard, the at least one processor performs several complex processing tasks. The at least one processor is communicably coupled to other components of the system device wirelessly and/or in a wired manner. In an example, the at least one processor may be implemented as a programmable digital signal processor (DSP). In another example, the at least one processor may be implemented via a cloud server that provides a cloud computing service.
Optionally, a given data source is implemented as at least one of: a data source of the system, an external data source. Herein, when the given data source is implemented as the data source of the system, data pertaining to the system is stored at the data source of the system and when the given data source is implemented as the external data source, data pertaining to the system is stored at the external data source.
Optionally, the input data comprises at least one of: historic data, real-time data, of the industrial process.
Optionally, the at least one data source generates control signals for controlling the industrial process in real-time.
Optionally, the at least one data source is implemented as at least one of: a virtual sensor, a physical sensor, a device.
Optionally, the at least one predictive domain model comprises at least one of: a material model, a ball charge model, a liner wear model, a dynamic charge model, inferential trajectory model, ball breakage model, new ball ejection model, a material processing model, a data processing model, a waste management model, a packaging model, an assembly model, a printing model.
Optionally, prior to the step of predicting the first set of output data, the at least one processor is configured to: - create the at least one predictive domain model using at least one machine learning algorithm; and - train the at least one predictive domain model using the input data. Optionally, when determining the at least one change required in the industrial process, the at least one processor is configured to: - determine at least one proposed change based on the at least one predicted state and the at least one required state; - simulate the at least one proposed change in the at least one digital twin of the industrial process to generate a simulated output state; - assess whether the simulated output state is similar to the at least one required state; and - when the simulated output state is similar to the at least one required state, implement the at least one proposed change as the at least one change required in the industrial process.
Optionally, the at least one processor is further configured to: - calculate an uncertainty of the at least one process action for implementing the at least one change; compare the uncertainty with a predefined uncertainty threshold; and - when the uncertainty is greater than the predefined uncertainty threshold, block at least one data entry from the first set of input data.
Optionally, the at least one processor is configured to employ adaptive data encryption and data obfuscation processing operations depending upon the at least one parameter of the industrial process.
Optionally, the at least one processor is further configured to send a notification to at least one device associated with an entity, wherein the notification is indicative of the at least one change, and wherein the entity performs the at least one change to optimize the industrial process.
Throughout the description and claims of this specification, the words "comprise" and "contain" and variations of the words, for example "comprising" and "comprises", mean "including but not limited to", and do not exclude other components, integers or steps. Moreover, the singular encompasses the plural unless the context otherwise requires: in particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.
Preferred features of each aspect of the invention may be as described in connection with any of the other aspects. Within the scope of this application, it is expressly intended that the various aspects, embodiments, examples and alternatives set out in the preceding paragraphs, in the claims and/or in the following description and drawings, and in particular the individual features thereof, may be taken independently or in any combination. That is, all embodiments and/or features of any embodiment can be combined in any way and/or combination, unless such features are incompatible.
BRIEF DESCRIPTION OF THE DRAWINGS
One or more embodiments of the invention will now be described, by way of example only, with reference to the following diagrams wherein: Figure 1 is a process flow depicting steps of a method for managing an industrial process, in accordance with an embodiment of the present disclosure; Figure 2 is a block diagram representing a system for managing an industrial process, in accordance with an embodiment of the present disclosure; and Figure 3 is an exemplary process flow depicting steps of a method for managing an industrial process, in accordance with another embodiment of the present disclosure.
DETAILED DESCRIPTION OF THE DRAWINGS
Referring to Figure 1, illustrated is a process flow depicting steps of a method for managing an industrial process, in accordance with an embodiment of the present disclosure. At step 102, information pertaining to the industrial process is obtained. At step 104, at least one digital twin of the industrial process is created, based at least on the information pertaining to the industrial process, wherein the at least one digital twin is dynamically updated while the industrial process is executed. At step 106, a first set of input data is received from at least one data source, wherein the first set of input data is indicative of at least one parameter in the industrial process. At step 108, a first set of output data is predicted by executing at least one predictive domain model that is pre-trained, wherein the first set of output data is predicted based on the first set of input data and the at least one digital twin, and wherein the first set of output data is representative of at least one predicted state of the industrial process. At step 110, the at least one predictive domain model that is pre-trained is executed using an addition of a first set of output data and the first set of input data to generate a second set of input data, wherein the second set of input data is an accurate representation of the at least one predicted state of the industrial process. At step 112, at least one change required in the industrial process is determined, based on the at least one predicted state and at least one required state of the industrial process. At step 114, at least one process action to implement the at least one change is initiated in response to the second set of input data for optimizing the industrial process.
The aforementioned steps are only illustrative and other alternatives can also be provided where one or more steps are added, one or more steps are removed, or one or more steps are provided in a different sequence without departing from the scope of the claims herein.
Referring to Figure 2, illustrated is a block diagram representing a system 200 for managing an industrial process, in accordance with an embodiment of the present disclosure. The system 200 comprises at least one data source (depicted as a data source 202) and at least one processor (depicted as a processor 204). The data source 202 and the processor 204 are communicably coupled.
It may be understood by a person skilled in the art that the Figure 2 is merely an example for sake of clarity, which should not unduly limit the scope of the claims herein. The person skilled in the art will recognize many variations, alternatives, and modifications of embodiments of the present disclosure.
Referring to Figure 3, illustrated is an exemplary process flow depicting steps of a method for managing an industrial process, in accordance with another embodiment of the present disclosure. At step 302, information pertaining to the industrial process is obtained. At step 304, at least one digital twin of the industrial process is created, based at least on the information pertaining to the industrial process, wherein the at least one digital twin is dynamically updated while the industrial process is executed. At step 306, a first set of input data is received from at least one data source, wherein the first set of input data is indicative of at least one parameter in the industrial process. At step 308, at least one predictive domain model is created using at least one machine learning algorithm. At step 310, the at least one predictive domain model is trained using the input data. At step 312, a first set of output data is predicted by executing the at least one predictive domain model that is pre-trained, wherein the first set of output data is predicted based on the first set of input data and the at least one digital twin, and wherein the first set of output data is representative of at least one predicted state of the industrial process. At step 314, the at least one predictive domain model that is pre-trained is executed using an addition of a first set of output data and the first set of input data, to generate a second set of input data, wherein the second set of input data is an accurate representation of the at least one predicted state of the industrial process. At step 316, at least one change required in the industrial process is determined, based on the at least one predicted state and at least one required state of the industrial process. Herein, at step 316a, at least one proposed change is determined based on the at least one predicted state and the at least one required state. At step 316b, the at least one proposed change is simulated in the at least one digital twin of the industrial process to generate a simulated output state. At step 316c, whether the simulated output state is similar to the at least one required state is assessed. At step 316d, when the simulated output state is similar to the at least one required state, the at least one proposed change is implemented as the at least one change required in the industrial process. At step 318, at least one process action to implement the at least one change is initiated in response to the second set of input data for optimizing the industrial process. At step 320, uncertainty of the at least one process action for implementing the at least one change is calculated. At step 322, the uncertainty is compared with a predefined uncertainty threshold. At step 324, when the uncertainty is greater than the predefined uncertainty threshold, at least one data entry is blocked from the first set of input data. At step 326, the at least one change is sent to at least one device associated with an entity, wherein the entity performs at least one change to optimize the industrial process.
The aforementioned steps are only illustrative and other alternatives can also be provided where one or more steps are added, one or more steps are removed, or one or more steps are provided in a different sequence without departing from the scope of the claims herein.
Claims (20)
- CLAIMSWhat is claimed is: 1. A method for managing an industrial process, the method comprising: - obtaining information pertaining to the industrial process; - creating at least one digital twin of the industrial process based at least on the information pertaining to the industrial process, wherein the at least one digital twin is dynamically updated while the industrial process is executed; - receiving a first set of input data from at least one data source, wherein the first set of input data is indicative of at least one parameter in the industrial process; - predicting a first set of output data by executing at least one predictive domain model that is pre-trained, wherein the first set of output data is predicted based on the first set of input data and the at least one digital twin, and wherein the first set of output data is representative of at least one predicted state of the industrial process; - executing the at least one predictive domain model that is pre-trained, using an addition of a first set of output data and the first set of input data, to generate a second set of input data, wherein the second set of input data is an accurate representation of the at least one predicted state of the industrial process; - determining at least one change required in the industrial process, based on the at least one predicted state and at least one required state of the industrial process; and - initiating, in response to the second set of input data, at least one process action to implement the at least one change, for optimizing the industrial process.
- 2. The method of claim 1, wherein input data comprises at least one of: historic data, real-time data, of the industrial process.
- 3. The method of claim 1 or 2, wherein the at least one data source generates control signals for controlling the industrial process in real-time.
- 4. The method of claim 1, 2 or 3, wherein the at least one data source is implemented as at least one of: a virtual sensor, a physical sensor, a device.
- 5. The method of any of the preceding claims, wherein the at least one predictive domain model comprises at least one of: a material model, a ball charge model, a liner wear model, a dynamic charge model, inferential trajectory model, ball breakage model, new ball ejection model, a material processing model, a data processing model, a waste management model, a packaging model, an assembly model, a printing model.
- 6. The method of any of the preceding claims, wherein prior to the step of predicting the first set of output data, the method comprises: - creating the at least one predictive domain model using at least one machine learning algorithm; and - training the at least one predictive domain model using the input data.
- 7. The method of any of the preceding claims, wherein when determining the at least one change required in the industrial process, the method is configured to: - determining at least one proposed change based on the at least one predicted state and the at least one required state; - simulating the at least one proposed change in the at least one digital twin of the industrial process to generate a simulated output state; - assessing whether the simulated output state is similar to the at least one required state; and - when the simulated output state is similar to the at least one required state, implementing the at least one proposed change as the at least one change required in the industrial process.
- 8. The method of any of the preceding claims, wherein the method further comprises: - calculating an uncertainty of the at least one process action for implementing the at least one change; - comparing the uncertainty with a predefined uncertainty threshold; and - when the uncertainty is greater than the predefined uncertainty threshold, blocking at least one data entry from the first set of input data.
- 9. The method of any of the preceding claims, wherein the method is configured to employ adaptive data encryption and data obfuscation processing operations depending upon the at least one parameter of the industrial process.
- 10. The method any of the preceding claims, wherein the method further comprises sending a notification, to at least one device associated with an entity, wherein the notification is indicative of the at least one change, and wherein the entity performs the at least one change to optimize the industrial process.
- 11. A system for managing an industrial process, the system comprising at least one data source and at least one processor, wherein the at least one data source and the at least one processor are communicably coupled, and wherein the at least one processor is configured to: - obtain information pertaining to the industrial process; - create at least one digital twin of the industrial process based at least on the information pertaining to the industrial process, wherein the at least one digital twin is dynamically updated while the industrial process is executed; - receive a first set of input data from the at least one data source, wherein the first set of input data is indicative of at least one parameter in the industrial process; - predict a first set of output data by executing at least one predictive domain model that is pre-trained, wherein the first set of output data is predicted based on the first set of input data and the at least one digital twin, and wherein the first set of output data is representative of at least one predicted state of the industrial process; - executing the at least one predictive domain model that is pre-trained, using an addition of a first set of output data and the first set of input data, to generate a second set of input data, wherein the second set of input data is an accurate representation of the at least one predicted state of the industrial process; - determine at least one change required in the industrial process, based on the at least one predicted state and at least one required state of the industrial process; and - initiate, in response to the second set of input data, at least one process action to implement the at least one change, for optimizing the industrial process.
- 12. The system of claim 11, wherein input data comprises at least one of: historic data, real-time data, of the industrial process.
- 13. The system of claim 11 or 12, wherein the at least one data source generates control signals for controlling the industrial process in real-time.
- 14. The system of claim 11, 12 or 13, wherein the at least one data source is implemented as at least one of: a virtual sensor, a physical sensor, a device.
- 15. The system of any of the claims 11-14, wherein the at least one predictive domain model comprises at least one of: a material model, a ball charge model, a liner wear model, a dynamic charge model, inferential trajectory model, ball breakage model, new ball ejection model, a material processing model, a data processing model, a waste management model, a packaging model, an assembly model, a printing model.
- 16. The system of any of the claims 11-15, wherein prior to the step of predicting the first set of output data, the at least one processor is configured to: - create the at least one predictive domain model using at least one machine learning algorithm; and - train the at least one predictive domain model using the input data.
- 17. The system of any of the claims 11-16, wherein when determining the at least one change required in the industrial process, the at least one processor is configured to: - determine at least one proposed change based on the at least one predicted state and the at least one required state; - simulate the at least one proposed change in the at least one digital twin of the industrial process to generate a simulated output state; - assess whether the simulated output state is similar to the at least one required state; and - when the simulated output state is similar to the at least one required state, implement the at least one proposed change as the at least one change required in the industrial process.
- 18. The system of any of the claims 11-17, wherein the at least one processor is further configured to: - calculate an uncertainty of the at least one process action for implementing the at least one change; - compare the uncertainty with a predefined uncertainty threshold; and - when the uncertainty is greater than the predefined uncertainty threshold, block at least one data entry from the first set of input data.
- 19. The system of any of the claims 11-18, wherein the at least one processor is configured to employ adaptive data encryption and data obfuscation processing operations depending upon the at least one parameter of the industrial process.
- 20. The system of any of the claims 11-19, wherein the at least one processor is further configured to send a notification to at least one device associated with an entity, wherein the notification is indicative of the at least one change, and wherein the entity performs the at least one change to optimize the industrial process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| GB2212522.3A GB2622004A (en) | 2022-08-30 | 2022-08-30 | Method and system for managing industrial processes |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| GB2212522.3A GB2622004A (en) | 2022-08-30 | 2022-08-30 | Method and system for managing industrial processes |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| GB202212522D0 GB202212522D0 (en) | 2022-10-12 |
| GB2622004A true GB2622004A (en) | 2024-03-06 |
Family
ID=83931668
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| GB2212522.3A Pending GB2622004A (en) | 2022-08-30 | 2022-08-30 | Method and system for managing industrial processes |
Country Status (1)
| Country | Link |
|---|---|
| GB (1) | GB2622004A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3974918A1 (en) * | 2020-09-24 | 2022-03-30 | ATLAS COPCO AIRPOWER, naamloze vennootschap | A method for controlling a compressor room and an apparatus thereof |
| EP4220313A1 (en) * | 2021-11-08 | 2023-08-02 | Rockwell Automation Technologies, Inc. | Generating digital twin systems for multiphysics systems |
-
2022
- 2022-08-30 GB GB2212522.3A patent/GB2622004A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3974918A1 (en) * | 2020-09-24 | 2022-03-30 | ATLAS COPCO AIRPOWER, naamloze vennootschap | A method for controlling a compressor room and an apparatus thereof |
| EP4220313A1 (en) * | 2021-11-08 | 2023-08-02 | Rockwell Automation Technologies, Inc. | Generating digital twin systems for multiphysics systems |
Also Published As
| Publication number | Publication date |
|---|---|
| GB202212522D0 (en) | 2022-10-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20220146993A1 (en) | Machine learning method and machine learning device for learning fault conditions, and fault prediction device and fault prediction system including the machine learning device | |
| Skordilis et al. | A deep reinforcement learning approach for real-time sensor-driven decision making and predictive analytics | |
| Rodríguez-Ramos et al. | An approach to fault diagnosis with online detection of novel faults using fuzzy clustering tools | |
| US20200150601A1 (en) | Solution for controlling a target system | |
| JP2016538645A (en) | Method and system for control based on artificial intelligence model of dynamic processes using stochastic factors | |
| US20210390160A1 (en) | Base analytics engine modeling for monitoring, diagnostics optimization and control | |
| CN113168139A (en) | Prediction for processes in an industrial plant | |
| US11644823B2 (en) | Automatic modeling for monitoring, diagnostics, optimization and control | |
| US10921755B2 (en) | Method and system for competence monitoring and contiguous learning for control | |
| EP3985460B1 (en) | Controller, control system and method to control a facility and to generate a control model by learning | |
| Balta et al. | Digital twin-based cyber-attack detection framework for cyber-physical manufacturing systems | |
| EP4083866A1 (en) | Information processing device, method, and program | |
| CN117639602A (en) | Self-adaptive motor running state adjusting method and system | |
| Amin et al. | Cyber security and beyond: Detecting malware and concept drift in AI-based sensor data streams using statistical techniques | |
| Ren et al. | A three-level hierachical framework for additive manufacturing | |
| US20190102352A1 (en) | Multi-engine modeling for monitoring, diagnostics, optimization and control | |
| CN116976862B (en) | Factory equipment informatization management system and method | |
| Sung et al. | Design-knowledge in learning plant dynamics for detecting process anomalies in water treatment plants | |
| GB2622004A (en) | Method and system for managing industrial processes | |
| Ashfahani et al. | Autonomous deep quality monitoring in streaming environments | |
| GB2465861A (en) | A reasoning inference making tool for recommending actions based on a hybridisation of a data driven model and knowledge based logic. | |
| Biczó et al. | Safe overfitting of boosted tree algorithm in heat transfer modeling | |
| Toubakh et al. | Advanced data mining approach for wind turbines fault prediction | |
| CN114722025A (en) | Data prediction method, device and equipment based on prediction model and storage medium | |
| D Marković-Petrović | A model for dynamic cyber security risk assessment in the industrial IoT environment |