Artificial intelligence kit for optimizing industrial plant design

Abstract

The invention relates to an artificial intelligence kit for optimizing industrial plant design. The artificial intelligence search engine integrates engineering knowledge and three-dimensional wiring space processing. A graphical user interface (GUI) prompts a user to input a start point and an end point corresponding to a specified type of conduit via an interactive three-dimensional model of a space in an industrial plant. An artificial intelligence search engine determines an optimal path from a plurality of three-dimensional path options corresponding to each conduit based on design constraints and / or objectives. The machine learning model modifies the best path option based on a previous user action in response to a previous best path option. The GUI displays the modified path options via an interactive three-dimensional model of the space. A natural language processor executes to: i) query engineering information, ii) operate an artificial intelligence search engine to generate design options, iii) deploy user selected options, or iv) adjust modified path options in response to user actions.

Description

[0001] Cross-reference to related cases

[0002] This application claims priority to U.S. Provisional Patent Application No. 63 / 627,414, filed on January 31, 2024, which is hereby incorporated herein by reference. Background Art

[0003] The conventional engineering design process for engineering, procurement, and construction projects is often plagued by slow and iterative workflows that are mainly manual and can significantly extend project durations. A key issue is the inefficiency of the design and delivery processes in engineering, procurement, and construction projects when building energy infrastructure. Today's laborious manual methods typically cannot fully explore complex design spaces, resulting in suboptimal solutions, extended project durations, and increased costs. Design engineers in engineering, procurement, and construction companies face increasingly complex and competing multi-objective and multi-disciplinary scenarios, while also being required to meet cost and sustainability targets. Many existing solutions mainly focus on conventional two-dimensional design because current three-dimensional design methods are slow, costly, and generally not ideal for the environment. In addition, conventional rule-based or deterministic systems or algorithms may miss the best solutions. Brief Description of the Drawings

[0004] Figure 1 A block diagram illustrating an example system of an artificial intelligence suite for optimizing industrial plant design according to an embodiment;

[0005] Figure 2 An example high-level architecture of an artificial intelligence suite for optimizing industrial plant design according to an embodiment;

[0006] Figure 3 An example flowchart architecture of an artificial intelligence suite for optimizing industrial plant design according to an embodiment;

[0007] Figure 4 An example graphical user interface of an artificial intelligence suite for optimizing industrial plant design according to an embodiment;

[0008] Figure 5 Is a flowchart illustrating a computer-implemented method of an artificial intelligence suite for optimizing industrial plant design according to an embodiment; and

[0009] Figure 6 Is a block diagram illustrating an example hardware device that can implement the subject matter. Detailed Description

[0010] Embodiments of the present disclosure provide an artificial intelligence suite for optimizing industrial plant design. The artificial intelligence search engine integrates engineering knowledge and three-dimensional wiring space processing. The graphical user interface prompts the user to input the start and end points corresponding to a specified type of conduit via an interactive three-dimensional model of the space in the industrial plant. A plurality of three-dimensional path options are generated from the start point to the end point corresponding to each of the conduits.

[0011] The artificial intelligence search engine determines the optimal path option from the plurality of three-dimensional path options corresponding to each of the conduits based on design constraints and / or design goals. The machine learning model modifies the optimal path option based on previous user actions in response to previous optimal path options. The graphical user interface displays the modified path option via the interactive three-dimensional model of the space in the industrial plant. The natural language processor is executed to: i) query engineering information, ii) operate the artificial intelligence search engine to generate design options, iii) deploy the user-selected option, or iv) adjust the modified path option in response to user actions.

[0012] For example, the artificial intelligence search engine integrates engineering knowledge and three-dimensional wiring space processing, which includes engineering knowledge and space processing of a heating, ventilation, and air conditioning utility plant. The graphical user interface, via the interactive three-dimensional model of the heating, ventilation, and air conditioning utility plant, prompts the user to input the start points of the chilled water supply pipe and the chilled water return pipe and the end points of the chilled water supply and return pipes. The three-dimensional path finding engine generates a plurality of three-dimensional path options for each of the chilled water supply and return pipes. The artificial intelligence search engine uses design constraints and design goals to determine the optimal path option for each of the chilled water supply and return pipes. The machine learning model modifies the optimal path option using previous user actions in response to previous optimal path options of the chilled water supply and return pipes by increasing the three-dimensional distance between some surface areas of the chilled water supply and return pipes and some surface areas of the previously wired superheated steam supply pipe and superheated steam return pipe.

[0013] The graphical user interface shows the increased three-dimensional distance between some surface areas of the chilled water supply and return pipes and some surface areas of the superheated steam supply and return pipes in a heating, ventilation, and air conditioning utility plant. The large language model provides the user with a technical explanation. Although the industrial engineering layout guidelines state that with proper insulation, the outer surfaces of the superheated steam supply and return pipes should not become hot enough to radiate sufficient heat to raise the temperature of the chilled water supply pipe or the chilled water return pipe, some industrial engineers have increased the minimum allowable three-dimensional distance between the surface areas of the chilled water supply and return pipes and the surface areas of the superheated steam supply and return pipes. After being satisfied with the technical explanation of the large language model, the user instructs the large language model to query engineering information to identify the nearest remaining three-dimensional distance between the surface areas of the chilled water supply and return pipes and the surface areas of the superheated steam supply and return pipes. Then the user instructs the large language model to instruct the artificial intelligence search engine to determine the maximum possible increase in the three-dimensional distance between the surfaces within the limits of the surrounding structure. Then the large language model enables the user to increase the three-dimensional distance between some surface areas of the chilled water supply and return pipes and some surface areas of the superheated steam supply and return pipes, and then deploys the two modified three-dimensional paths of the chilled water supply and return pipes.

[0014] Various embodiments and aspects of the present disclosure will be described with reference to the details discussed below, and the drawings will illustrate the various embodiments. The following description and the drawings are illustrative of the present disclosure and should not be construed as limiting the present disclosure. Many specific details are described to provide a thorough understanding of the various embodiments of the present disclosure. However, in some cases, well-known or conventional details are not described in order to provide a concise discussion of the embodiments of the present disclosure.

[0015] Although these embodiments are described in sufficient detail to enable those skilled in the art to practice the disclosed embodiments, it should be understood that these examples are not restrictive, so other embodiments may be used and changes may be made without departing from their spirit and scope. For example, the operations of the methods shown and described herein need not be performed in the order indicated and may be performed in parallel. It should also be understood that these methods may include more or fewer operations than indicated. In some embodiments, the operations described herein as separate operations may be combined. Conversely, an operation that may be described herein as a single operation may be implemented in multiple operations.

[0016] The phrase "an embodiment" or "embodiment" or "some embodiments" mentioned in the specification means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the present disclosure. The phrase "embodiment" or "the embodiment" that appears throughout the specification does not necessarily refer to the same embodiment.

[0017] Figure 1The block diagram of an example system 100 of an artificial intelligence suite for optimizing the design of industrial plants according to an embodiment is illustrated. As Figure 1 shown, the system 100 may illustrate a cloud computing environment where data, applications, services, and other resources are stored and delivered through a shared data center and appear as a single access point for users. The system 100 may also represent any other type of distributed computer network environment where servers control the storage and distribution of resources and services for different client users.

[0018] In an embodiment, the system 100 represents a cloud computing system that includes a first client 102, a second client 104, a third client 106, a fourth client 108, a server 110, and an optional cloud computing environment 112 that may be provided by a hosting company. The clients 102 - 108, the server 110, and the cloud computing environment 112 communicate via a network 114. Although Figure 1 the first client 102 is depicted as a laptop computer 102, the second client 104 is depicted as a desktop computer 104, the third client 106 is depicted as a smart phone 106, and the fourth client 108 is depicted as a server, each of the system components 102 - 110 can be any type of computer system and each can be substantially similar to Figure 6 the hardware device 600 depicted herein and described below.

[0019] The server 110 may host and execute an artificial intelligence design system 116 that generates an industrial engineering layout design, which can be accessed via a graphical user interface 118, as Figure 1 depicted, and / or reside on any of the clients 102 - 108. Although Figure 1 it is depicted that the entire artificial intelligence design system 116 resides completely on the server 110, any or all of the artificial intelligence design system 116 can reside completely on the clients 102 - 108, completely on the cloud computing environment 112, or in any combination of partially on the clients 102 - 108, partially on the server 110, partially on the cloud computing environment 112, and / or partially on Figure 1 another server not depicted herein. Figure 1Depicts a system 100 with four clients 102 - 108, a server 110, a cloud computing environment 112, a network 114, an artificial intelligence design system 116, and a graphical user interface 118. However, system 100 can include any number of clients 102 - 108, any number of servers 110, any number of cloud computing environments 112, any number of networks 114, any number of artificial intelligence design systems 116, and any number of graphical user interfaces 118. The artificial intelligence design system 116 includes an artificial intelligence search engine 120, a machine learning module 122, and a large language model 124, each of which will be described in more detail below.

[0020] Figure 2 Depicts a circular diagram that describes the high - level architecture 200 of an automated and interactive artificial intelligence design system 116 for generating industrial engineering layout designs. The artificial intelligence design system 116 includes three operating layers in the high - level architecture 200. The first operating layer is the artificial intelligence search engine 120, which is activated whenever a human designer uses the artificial intelligence search assistant to build the best three - dimensional design model faster and more easily. The second operating layer is the machine learning module 122, which combines a number of machine learning models that provide a predictive design solution layer, where the machine learning models as a service are deployed in the background by the human designer and they learn, adapt, and dynamically suggest design elements. The third operating layer is the natural language processor 126, in the form of a bidirectional large language model, which is part of the conversational solution layer, where the human designer searches for one - dimensional, two - dimensional, and / or three - dimensional engineering information and performs discipline - specific tasks, such as building a three - dimensional model via specifications in statement format.

[0021] Figure 3Depicts a flowchart architecture 300 that highlights where data comes from, how data from different sources and processes is combined and passed to the algorithms of the artificial intelligence search engine 120 using input channels, and how the machine learning module 122 and the large language model 124 for search and query techniques are used to further enrich this data for different scenarios. The difference in the artificial intelligence design system 116 is how data is ingested, processed, and then converted into executable three-dimensional design element options through the artificial intelligence algorithms of the artificial intelligence search engine 120 that focus on exploration and development. Different from conventional design systems, design engineers must provide a small amount of upfront information via a general interface engine, specify boundaries, constraints, and goals (which can be called directives), and continue to pass them to the artificial intelligence search engine 120, which belongs to the sub-symbolic paradigm and requires little historical information of the problem statement but still quickly generates inherently out-of-the-box conflict-free design wiring options. The artificial intelligence search engine 120 uses evolutionary search and optimization algorithms to navigate the vast and complex three-dimensional design space.

[0022] The evolutionary search and optimization using sub-symbolic artificial intelligence algorithms are significant features. Different from rule-based or deterministic systems, these algorithms enable the artificial intelligence search engine 120 to iteratively explore the vast design space and discover solution options that may not be found by conventional engineering methods.

[0023] The key differences and advantages of the artificial intelligence design system 116 over alternative methods lie in its evolutionary leap. The artificial intelligence design system 116 makes full use of evolutionary search and optimization algorithms, thus differentiating it from deterministic methods. The sub-symbolic artificial intelligence foundation enables the artificial intelligence search engine 120 to iteratively explore and refine design options, thereby improving adaptability and discovering the best solutions that conventional rule-based systems may miss. The integrated operation of a hybrid design space exploration / development artificial intelligence search strategy for three-dimensional space wiring problems. The hybrid exploration / development artificial intelligence search strategy combines user requirements with engineering knowledge to provide a set of cohesive deployable results on demand.

[0024] The sub-symbolic artificial intelligence search engine 120 can be enhanced using the machine learning module 122, which combines and utilizes multiple machine learning models to more accurately and relevantly formulate questions and generate more relevant outputs via reliability, constructability, and feasibility checks. The machine learning module 122 can continuously self-learn, adjust, and better examine the output of the generation layer through a positive feedback loop. Machine learning integration plays a key role in enhancing the decision-making capabilities of the enhanced artificial intelligence design system 116. This data-driven approach not only simplifies the design process but also ensures that decisions are based on real-world outcomes, thereby improving the overall efficiency and effectiveness of design projects.

[0025] Incorporating machine learning into decision-making represents a paradigm shift. The artificial intelligence design system 116 no longer solely relies on deterministic algorithms but can learn from historical project data. This adaptive learning enables the artificial intelligence design system 116 to make informed design choices, thereby improving accuracy and optimizing solutions based on real-world outcomes.

[0026] Incorporating machine learning facilitates data-driven decision-making. The machine learning module 122 learns from historical project data, enabling it to predict the best design choices, adapt to changing project requirements, and continuously improve its decision-making accuracy. This dynamic learning ability differentiates it from static algorithms and enhances the artificial intelligence design system 116's ability to address the evolving challenges in energy infrastructure design projects.

[0027] The machine learning module 122 is capable of integrating reference data sets to learn and guide the artificial intelligence algorithm search process and can build usage patterns based on the accessed data and actions taken by the user in the form of feedback. Subsequently, it will also automatically provide the ability to build human selection patterns in the form of a feedback loop or historical reference data sets to guide the artificial intelligence search engine 120 in subsequent processes in context, thereby making design decisions faster and better.

[0028] Since both the generative layer and the predictive layer can be deployed for any engineering discipline, they can be extended across all these disciplines to work in parallel and communicate seamlessly, providing early warnings or influencing considerations from engineers in one discipline to those in another. The conversational AI layer of the bidirectional large language model 124 then integrates and wraps these two underlying layers to form an easy-to-use prompt-based natural language tool for rapid engineering search, aggregation, and execution of workflows, all through the same graphical user interface 118 for the entire project team. Using the large language model 124 to perform iterative design optimization helps enable seamless communication and collaboration in natural language among different stakeholders. Multidisciplinary engineers and other stakeholders can make all these requests for guidance, search, and rapid aggregation of information or construct and edit 3D design elements via prompts in a real-time collaborative environment.

[0029] The large language model 124 is a collaborative tool that introduces a new dimension to stakeholder participation. Different from conventional design systems, this feature allows multidisciplinary engineers and domain experts to communicate and contribute using natural language. This enhances multidisciplinary collaboration and promotes a more inclusive and efficient 3D design process.

[0030] Figure 4 An example graphical user interface 400 of an AI suite for optimizing industrial plant design according to an embodiment is illustrated. As a collaborative tool, the large language model 124 abandons the cumbersome interface and introduces a novel and inclusive dimension to stakeholder participation using natural language. This promotes effective communication and knowledge transfer between multidisciplinary engineers and domain experts. Different from alternatives that may lack a natural language interface, this feature simplifies collaboration and ensures a multidisciplinary approach to design decisions.

[0031] The AI design system 116 is a generative AI-driven engineering design system for accelerating engineering, procurement, and construction projects in the energy infrastructure. The conventional engineering design process for engineering, procurement, and construction projects is often plagued by slow and iterative workflows that are mainly manual and can significantly extend project duration. To eliminate this key bottleneck, the AI design system 116 aims to transform the engineering design phase in engineering, procurement, and construction projects by developing and deploying an AI search engine 120 that can quickly, automatically, and interactively create 3D design model options (such as pipes, structures, equipment, electrical, heating, ventilation, and air conditioning, etc.) for users to select and implement. The AI design system 116 replaces today's conventional laborious human-driven design process with an AI-driven but human-selected future design process, addressing various challenges in the design and development of certain energy infrastructure.

[0032] A key issue is the inefficiency of the conventional design and delivery processes for engineering, procurement, and construction projects in building energy infrastructure. Conventional human-driven manual methods often lack the ability to comprehensively explore intricate design spaces, leading to suboptimal solutions, extended project durations, and increased costs. The AI Design System 116 is an AI-driven 3D generative design system that aims to revolutionize this paradigm by leveraging sub-symbolic AI search algorithms, machine learning, and large language models. Faster and cheaper energy infrastructure engineering, procurement, and construction projects mean lower energy costs, thus enhancing economic competitiveness and household safety. Efficiency is crucial in the transition to sustainable energy, and the AI Design System 116 accelerates decision-making, optimizes designs, and streamlines project delivery.

[0033] By prioritizing sustainability and integrating environmentally friendly practices, the AI Design System 116 aligns with the commitment to reducing carbon emissions, achieving renewable energy targets, and advancing environmental responsibility. The significance of this AI Design System 116 lies in its potential to reshape how energy infrastructure design is conceptualized and realized for sustainability. By focusing on sustainability and providing multi-dimensional generative design, the AI Design System 116 represents a leap forward from conventional AI-driven design systems. A framework for sustainable practices can also be incorporated, which guides the optimization system to prioritize design choices with reduced carbon footprints.

[0034] Integrating 3D generative design represents a significant departure from conventional 2D-centric approaches. By extending generative design to 3D, the AI Design System 116 enhances visualization and facilitates a more detailed exploration of spatial configurations. Conventional 3D design methods are slow, costly, and often suboptimal for the environment.

[0035] The cutting-edge AI Design System 116 addresses this challenge by making sustainability its core and transforming the way it designs processes or power plants. The emphasis on 3D generative design is groundbreaking. Many design systems focus primarily on 2D design, while the AI Design System 116 concentrates generative design capabilities in the 3D domain. This not only enhances visualization but also enables the exploration of complex spatial configurations, facilitating a more comprehensive understanding of sustainable design options. Since the AI Design System 116 consists of multiple layers of technology, embedded generative design, and a cloud-based large language model, ensuring modular design will allow for flexible implementation and adaptation to different user configurations, thus meeting a wider range of customer needs.

[0036] By transforming engineering much earlier in the design phase, the AI Design System 116 aims to maximize the productivity, efficiency, and environmental awareness of energy infrastructure projects. At the core of the AI Design System 116 is the incorporation of a cutting-edge 3D generative design system that brings together sub-symbolic AI algorithms, machine learning, and large language models, revolutionizing the design process of energy infrastructure. Overall, the AI Design System 116 can accelerate the delivery of engineering, procurement, and construction projects by significantly reducing design iteration time and optimizing resource allocation. The AI Design System 116 is innovative both in its underlying AI and its applications, setting it apart from existing products in the field.

[0037] The innovative aspects lie in its multi-level and coordinated AI technology suite and its revolutionary application of faster building and deployment of 3D designs. While existing design systems may utilize some of these elements, none offer the same level of comprehensive integration. In summary, the AI Design System 116 stands out for its integration of multiple AI technologies. By embracing innovation in design concepts and technical foundations, this AI Design System 116 provides a holistic, efficient, and superior alternative to conventional design methods.

[0038] An intelligent generative design search and optimization method is disclosed, which automates the design process of 3D wiring elements for industrial plant design. The method includes encoding engineering knowledge, 3D wiring space processing, requirement consideration, specifying design goals and design constraints for a given 3D design search problem, and subsequently automatically supplying relevant design variants to the designer, which are sorted contextually and interactively before deployment. This innovative AI Design System 116 has multiple advantages: (1) providing end-to-end automatic wiring of 3D elements on demand, generating optimal (constructible, always conflict-free, and sustainable) wiring options; (2) enhancing the designer's ability by enabling intelligent designer-as-a-service, which reduces the manual effort required, making collaboration among multidisciplinary 3D designers easier, thereby reducing costs, time scales, and commercial risks. The AI Design System 116 integrates engineering knowledge and practical rules and provides the ability to visualize any results of the output before deploying these results.

[0039] The AI Design System 116 also provides the ability to rank the output results based on metric goals before deploying these results, as well as the ability to adjust design goals and design constraints on demand to improve the quality of the output results. Different aspects of the AI Design System 116 operate in concert and provide the unique ability to specify and adjust design problems on demand, quickly search the design space, iteratively generate wiring options, rank and visualize these options using relevant information supplied by the user before acceptance.

[0040] Figure 5 is a flowchart of a computer-implemented method of an artificial intelligence suite for optimizing the design of an industrial plant according to an embodiment. Flowchart 500 depicts method actions that are illustrated as flowchart blocks for certain actions involved in and / or between system elements 102 - 124 Figure 1 of the system.

[0041] The artificial intelligence search engine integrates engineering knowledge and three-dimensional routing space processing, block 502. The system prepares the artificial intelligence search engine to assist in generating the design of an industrial plant. For example, but not limited to, this can include the artificial intelligence search engine 120 integrating engineering knowledge and three-dimensional routing space processing, which includes engineering knowledge and space processing of heating, ventilation, and air conditioning utility plants.

[0042] The artificial intelligence search engine can be a computer software system that can retrieve information about the layout of an industrial plant. Engineering knowledge can be facts and information about the branch of science and technology involved in the design, construction, and use of engines, machines, and structures. Three-dimensional routing space processing can be how paths are presented through an area that appears to have height, length, and width dimensions.

[0043] Engineering knowledge and three-dimensional routing space processing can include rules of practice, requirements considerations, design goals, design constraints, cost metrics, and / or sustainability metrics. For example, rules of practice can be a series of if-then-else questions that a user must apply to each individual conduit that is to be routed through a three-dimensional model of the space in an industrial plant.

[0044] Rules of practice can be the idea or method of the actual application or use of principles that regulate behavior within a particular activity or field. Requirements considerations can be the careful consideration of what is needed or wanted. Design goals can be a plan to achieve production to show the appearance and function of an object before it is built or manufactured.

[0045] Design constraints can be limitations or restrictions on a plan to achieve production to show the appearance and function of an object before it is built or manufactured. Cost metrics can be the estimated price of a goal or result to be achieved. Sustainability metrics can be the ability of a goal or result to be achieved to be maintained at a certain rate.

[0046] After the artificial intelligence search engine integrates three-dimensional routing space processing, the graphical user interface prompts the user to input the start and end points corresponding to a specified type of conduit via an interactive three-dimensional model of the space in the industrial plant, block 504. The system prompts the user to provide information for initiating the automated design of the industrial plant layout. By way of example, but not limited to, this can include the graphical user interface 118 via such as Figure 4An interactive three-dimensional model of a heating, ventilation, and air conditioning (HVAC) utility plant drawn within a graphical user interface prompts a user to input starting points of chilled water supply pipes and chilled water return pipes, and to input ending points of the chilled water supply pipes and the chilled water return pipes.

[0047] The graphical user interface can be a computer monitor for human-computer system interaction, particularly an input device and software for human interaction. The user can be a person interacting with the computer. The starting point can be the beginning of a position in a specific location, place, or area. The ending point can be the final part of a position in a specific location, place, or area. A conduit can be a passage for conveying something.

[0048] The specified type can be a clearly identified category of things having common characteristics. The interactive three-dimensional model can be a computer representation of a proposed structure that appears to have length, width, and depth and allows two-way information flow between a computer and a computer user. Space can be an area having the dimensions of height, length, and width. An industrial plant can be a factory or a similar place for manufacturing and / or distribution.

[0049] The conduit can be a passage of an appropriate type for conveying water, liquid, gas, electricity, heat, and / or air conditioning. For example, the conduit can be an electric cable that supplies power to an air conditioning unit, or a ventilation duct that supplies air conditioning to people in an office building. Type can be a category or classification. Water can be a colorless, transparent, and odorless liquid. Liquid can be a substance that flows freely but has a constant volume.

[0050] Gas can be a substance in a state where it will freely expand to fill an entire container, having no fixed shape (unlike a solid) and no fixed volume (unlike a liquid). Electricity can be a form of energy generated by the presence of charged particles. Heat can be the property of being hot, or high temperature. Air conditioning can be a system for controlling humidity, ventilation, and temperature in a building, typically used to maintain a cool atmosphere under warm conditions.

[0051] After inputting starting and ending points corresponding to conduits of a specified type via the interactive three-dimensional model, multiple three-dimensional path options are generated from the starting point to the ending point corresponding to each conduit, block 506. The system iteratively generates path options for a specific type of conduit. In an embodiment, this can include a three-dimensional path finding engine that generates multiple three-dimensional path options for each chilled water supply and return pipe. The three-dimensional path options can be wirings that are selected or can be selected and appear to have length, width, and depth.

[0052] Generating path options for each conduit may include organizing the path options into several compatible groups of path options, where each group is ranked according to at least one of at least one design constraint and / or at least one design goal. For example, when routing four chilled water supply and return pipes from a heating, ventilation, and air conditioning utility plant, the artificial intelligence search engine 120 applies the highest weight to the design constraint and the second highest weight to the design goal to generate a set of four best path options, which are ranked first based on the design constraint and second based on the design goal. The artificial intelligence search engine 120 may also apply the second highest weight to the design constraint and the highest weight to the design goal to generate a slightly different set of six best path options, which are ranked first based on the design goal and second based on the design constraint. In another example, the artificial intelligence search engine 120 may use only one design constraint to generate a set of four best path options, which are ranked first based on the design constraint. In yet another example, the artificial intelligence search engine 120 may use only one design goal to generate a set of four best path options, which are ranked first based on the design goal. A compatible group may be a group or collection of things that are of the same kind, similar to each other, or commonly occur together and can exist or occur together without conflict.

[0053] After generating a plurality of three-dimensional path options from a starting point to an end point corresponding to each conduit, the artificial intelligence search engine determines the best path option from the plurality of three-dimensional path options corresponding to each conduit based on design constraints and / or design goals, block 508. The system uses the design goals and constraints of the conduit to identify the best path option. For example, but not limited to, this may include the artificial intelligence search engine 120 using design constraints and design goals to determine the best path option for each chilled water supply and return pipe. The best path option may be the best or most favorable routing that is selected or can be selected.

[0054] Design constraints and / or design goals for the conduit can be input via at least one of a graphical user interface or a natural language processor, and the best path options are constructible, sustainable based on layout options, and inherently conflict-free based on sorting the handling of the conduit. For example, the large language model 124 responds to the user selecting the start and end points of the chilled water supply and return pipes, and then asks the user whether they want to continue using the same design constraints and design goals selected by the previous user for the chilled water supply and return pipes, or update the design constraints and / or design goals for the chilled water supply and return pipes. In another example, the best path options are inherently conflict-free because the three-dimensional pathfinding engine sequentially identifies the paths of each conduit such that after the three-dimensional pathfinding engine identifies the first path option for the first conduit, the three-dimensional pathfinding engine will not consider any part of the first path option of the first conduit as a potential location for any part of the path options of the second, third, or fourth conduits.

[0055] After the three-dimensional pathfinding engine identifies the first path option for the second conduit, the three-dimensional pathfinding engine will not consider any part of the path options of the first or second conduits as a potential location for any part of the path options of the third or fourth conduits. The three-dimensional pathfinding engine continues this process until the path options for all conduits have been identified. If a conduit requires more path options, then the three-dimensional pathfinding engine repeats this process from the beginning until a sufficient number of path options for all conduits have been identified.

[0056] Constructibility can refer to the ability to build, assemble, or combine, and sustainability can refer to the ability to maintain at a certain rate or level. Layout options can refer to the way the parts of something are arranged, and the way they are selected or can be selected. Conflict-free can refer to the absence or non-existence of opposition.

[0057] After the artificial intelligence search engine determines the best path options from the multiple three-dimensional path options corresponding to each conduit using the design constraints and / or design goals, the machine learning model modifies the best path options based on previous user actions in response to the previous best path options, block 510. The system learns how to improve some of the best path options. By way of example, but not limited to, this can include the machine learning model 122 using previous user actions in response to the previous best path options for the chilled water supply and return pipes to modify the best path options by increasing the three-dimensional distance between some surface areas on the chilled water supply and return pipes and some surface areas on the superheated steam supply and return pipes.

[0058] The artificial intelligence design system 116 follows the required three-dimensional distance between any surface areas on the chilled water supply and return pipes and any surface areas on the superheated steam supply and return pipes in the published industrial engineering guidelines to prevent the hot superheated steam supply and return pipes from inadvertently raising the temperature of the much cooler chilled water supply and return pipes. However, some industrial engineers who used the artificial intelligence design system 116 modified their displayed optimal path options to increase the three-dimensional distance between some surface areas on the chilled water supply and return pipes and some surface areas on the superheated steam supply and return pipes. The machine learning model 122 infers that these industrial engineers increased the three-dimensional distance between some surface areas on the chilled water supply and return pipes and some surface areas on the superheated steam supply and return pipes as a precaution in case the insulation on the pipes might be insufficient to prevent the superheated steam supply and return pipes from raising the temperature of the chilled water supply and return pipes. Such learning increases the likelihood that the machine learning model 122 will make similar modifications to the optimal path options in the future. A previous user action can be the physical activities of a person when interacting with a computer in the past. A previous optimal path option can be the wiring that was once the best, or most advantageous, and has been selected or could have been selected.

[0059] After the machine learning model modifies the optimal path option, based on the previous user actions in response to the previous optimal path option, the graphical user interface displays the modified path option, block 512, via an interactive three-dimensional model of the space in the industrial plant. The system displays any modifications made by the machine learning model. In an embodiment, this can include the graphical user interface 118 displaying the increased three-dimensional distance between the surface areas on the chilled water supply and return pipes and some surface areas on the superheated steam supply and return pipes.

[0060] After a modified path option has been displayed via the interactive three-dimensional model of the space in the industrial plant, the natural language processor executes to: i) query engineering information, ii) operate an artificial intelligence search engine to generate design options, iii) deploy the user-selected option, or iv) adjust the modified path option in response to a user action, block 514. The system enables the user to select from several final actions, which include querying engineering information, using artificial intelligence, and / or using other options to adjust the path option, and then deploying the path option in the industrial plant. For example, but not limited to, this can include the large language model 124 enabling the user to request engineering information to determine the location of the surrounding structures that may limit the increase in the three-dimensional distance between the surfaces of the chilled water and superheated steam pipes, and instructing the artificial intelligence search engine to determine the maximum possible increase in the three-dimensional distance between the surfaces within the limits of the surrounding structures. Then, the large language model 124 enables the user to further increase the three-dimensional distance between some surface areas on the chilled water supply and return pipes and some surface areas on the superheated steam supply and return pipes beyond any increase that the machine learning model has initiated, and then deploy the adjusted and modified three-dimensional path option. The adjustment can be a modification or an adjustment. The user action can be the activity of a person interacting with the computer. The adjustment can be a modification or an adjustment. The deployment can be to enable an effective action, or to utilize.

[0061] The machine learning model 122 can learn from a reference dataset as part of the training of the machine learning model 122 to guide the artificial intelligence search process, learn from the natural language processor to adjust the modified path option, and / or generate additional relevant paths via reliability, constructability, and / or feasibility checks. For example, the machine learning model 122 records that the user listened to the large language model 124 explain the reason for the increase in the three-dimensional distance between some surface areas on the chilled water supply and return pipes and some surface areas on the superheated steam supply and return pipes. Next, the user instructed the large language model 124 to identify the remaining surface areas with the closest three-dimensional distance between any surface areas on the chilled water supply and return pipes and any surface areas on the superheated steam supply and return pipes. Then, the user adjusted the modified path option to further increase the three-dimensional distance between some points on the chilled water supply and return pipes and some points on the superheated steam supply and return pipes. Such learning increases the likelihood that the machine learning model 122 will make similar modifications to the optimal path option in the future. The reference dataset can be a group of structured information held in a computer and used to determine something. The modified path option can be a revised route that has been selected or can be selected. The relevant path can be a route suitable for consideration. The reliability check can be a check to test or determine credibility. The constructability check can be a check to test or determine the ease, efficiency, and environmental friendliness of constructing a structure. The feasibility check is a check to test or determine the state or degree of ease or convenience of completion.

[0062] Adjustments made via a natural language processor may include adjusting design goals and / or design constraints. The natural language processor may be a bidirectional large language model that searches a three-dimensional model associated with engineering information and performs discipline-specific tasks. For example, the large language model 124 has an extended vocabulary to communicate with engineers in all disciplines about highly specialized technical topics. Thus, the large language model 124 immediately understands and interprets the role of superheated steam in terms of increasing the three-dimensional distance of some surface areas of the chilled water supply and return pipes relative to some surface areas of the superheated steam supply and return pipes.

[0063] However, an electrical engineer responsible for wiring a cable to a solenoid valve motor that will operate the valves of the superheated steam supply and return pipes may not immediately recognize the reference to superheated steam. The bidirectional large language model may be an artificial neural network that enables general communication with humans. The discipline-specific tasks may be the workload associated with the engineering field.

[0064] Although Figure 5 the blocks 502-514 are depicted as occurring in a particular order, the blocks 502-514 may occur in other orders. In other embodiments, each of the blocks 502-514 may also be executed in combination with other blocks and / or some blocks may be divided into a different set of blocks.

[0065] Exemplary hardware devices that may implement the present subject matter will be described. Those of ordinary skill in the art will recognize that Figure 6 the elements shown in Figure 6 , an exemplary system for implementing the subject matter disclosed herein includes a hardware device 600 including a processing unit 602, a memory 604, a storage device 606, a data input module 608, a display adapter 610, a communication interface 612, and a bus 614 that couples the elements 604-612 to the processing unit 602.

[0066] The bus 614 may include any type of bus architecture. Examples include a memory bus, a peripheral bus, a local bus, etc. The processing unit 602 is an instruction execution machine, device, or apparatus and may include a microprocessor, a digital signal processor, a graphics processing unit, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), etc. The processing unit 602 may be configured to execute program instructions stored in the memory 604 and / or the storage device 606 and / or received via the data input module 608.

[0067] The memory 604 may include a read-only memory (ROM) 616 and a random access memory (RAM) 618. The memory 604 may be configured to store program instructions and data during the operation of the hardware device 600. In various embodiments, the memory 604 may include any of a variety of memory technologies, such as static random access memory (SRAM) or dynamic RAM (DRAM), including variants such as, for example, double data rate synchronous DRAM (DDR SDRAM), error-correcting code synchronous DRAM (ECC SDRAM), or RAMBUS DRAM (RDRAM).

[0068] The memory 604 may also include non-volatile memory technologies, such as non-volatile flash RAM (NVRAM) or ROM. In some embodiments, it is contemplated that the memory 604 may include a combination of technologies such as those described above, as well as other technologies not specifically mentioned. When the subject is implemented in a computer system, the basic input / output system (BIOS) 620 is stored in the ROM 616 and contains basic routines that help transfer information between elements within the computer system during startup.

[0069] The storage device 606 may include a flash data storage device for reading and writing flash memory; a hard disk drive for reading and writing hard disks; a disk drive for reading or writing removable disks; and / or an optical disk drive for reading or writing removable optical disks, such as CD ROM, DVD, or other optical media. The drives and their associated computer-readable media provide non-volatile storage of computer-readable instructions, data structures, program modules, and other data for the hardware device 600.

[0070] Note that the methods described herein can be implemented as executable instructions stored on a computer-readable medium for use by or in conjunction with an instruction-executing machine, apparatus, or device such as a computer-based or processor-containing machine, apparatus, or device. Those skilled in the art will recognize that for some embodiments, other types of computer-readable media that can store computer-accessible data, such as magnetic tape, flash memory cards, digital video discs, Bernoulli boxes, RAM, ROM, etc., can also be used in exemplary operating environments. As used herein, "computer-readable medium" can include one or more any suitable media for storing executable instructions of a computer program in one or more of electronic, magnetic, optical, and electromagnetic formats such that the instruction-executing machine, system, apparatus, or device can read (or obtain) the instructions from the computer-readable medium and execute the instructions for performing the method. A non-exhaustive list of conventional exemplary computer-readable media includes: portable computer disks; RAM; ROM; erasable programmable read-only memory (EPROM or flash memory); optical storage devices, including portable compact discs (CDs), portable digital video discs (DVDs), high-definition DVDs (HD-DVD TM ), Blu-ray discs; and the like.

[0071] A plurality of program modules can be stored on the storage device 606, ROM 616, or RAM 618, including an operating system 622, one or more application programs 626, program data 626, and other program modules 628. A user can input commands and information into the hardware device 600 through the data input module 608. The data input module 608 can include mechanisms such as a keyboard, a touch screen, a pointing device, etc.

[0072] Other external input devices (not shown) are connected to the hardware device 600 via the external data input interface 610. By way of example and not limitation, the external input devices can include a microphone, a joystick, a gamepad, a satellite antenna, a scanner, etc. In some embodiments, the external input devices can include video or audio input devices such as a video camera, a still camera, etc. The data input module 608 can be configured to receive inputs from one or more users of the hardware device 600 and pass these inputs to the processing unit 602 and / or the memory 604 via the bus 614.

[0073] The display 612 is also connected to the bus 614 via a display adapter 610. The display 612 can be configured to display the output of the hardware device 600 to one or more users. In some embodiments, a given device (such as a touch screen) can, for example, be used as both the data input module 608 and the display 612 simultaneously. An external display device can also be connected to the bus 614 via an external display interface 634. Other peripheral output devices (such as speakers and printers) not shown can be connected to the hardware device 600.

[0074] The hardware device 600 can operate in a networked environment using a logical connection with one or more remote nodes (not shown) via the communication interface 612. The remote nodes can be another computer, server, router, peer device, or other common network node, and typically include many or all of the elements described above that are associated with the hardware device 600. The communication interface 612 can interface with a wireless network and / or a wired network. Examples of wireless networks include, for example, Bluetooth networks, wireless personal area networks, wireless 802.21 local area networks (LANs), and / or wireless telephone networks (such as cellular, PCS, or GSM networks).

[0075] Examples of wired networks include, for example, LANs, fiber optic networks, wired personal area networks, telephone networks, and / or wide area networks (WANs). Such networked environments are common in intranets, the Internet, offices, enterprise-wide computer networks, etc. In some embodiments, the communication interface 612 can include logic configured to support direct memory access (DMA) transfers between the memory 604 and other devices.

[0076] In a networked environment, program modules or portions thereof depicted with respect to the hardware device 600 can be stored in a remote storage device, such as, for example, on a server. It will be appreciated that other hardware and / or software can be used to establish a communication link between the hardware device 600 and other devices.

[0077] It should be understood that Figure 6 the arrangement of the hardware device 600 shown is only one possible implementation, and other arrangements are also feasible. It should also be understood that the various system components (and parts) defined by the claims, described below, and shown in the various block diagrams represent logical components configured to perform the functions described herein. For example, one or more of these system components (and parts) can be implemented, in whole or in part, by at least some of the components shown in the arrangement of the hardware device 600.

[0078] In addition, while at least one of these components is at least partially implemented as an electronic hardware component and thus constitutes a machine, other components may be implemented in software, hardware, or a combination of software and hardware. More particularly, at least one component defined by the claims is at least partially implemented as an electronic hardware component, such as an instruction-executing machine (e.g., a processor-based or processor-containing machine) and / or a dedicated circuit or circuitry (e.g., discrete logic gates interconnected to perform a dedicated function), such as Figure 6 those shown in

[0079] Other components may be implemented in software, hardware, or a combination of software and hardware. In addition, some or all of these other components may be combined, some may be entirely omitted, and additional components may be added while still implementing the functions described herein. Thus, the subject matter described herein may be implemented in many different variations, and all such variations are considered to be within the scope of the claims.

[0080] In the foregoing description, unless otherwise indicated, the subject matter is described with reference to symbolic representations of acts and operations performed by one or more devices. From this, it can be understood that these acts and operations, sometimes referred to as being performed by a computer, include the manipulation of data in structured form by a processing unit. Such manipulation transforms the data or maintains its position in the computer's memory system, thereby reconfiguring or otherwise changing the operation of the device in a manner well known to those skilled in the art. The physical location of the data that maintains the data is a memory having specific properties defined by the data format. However, while the subject matter is described in a context, this is not meant to be limiting, as those skilled in the art will recognize that the various acts and operations described hereinafter may also be implemented in hardware.

[0081] To facilitate understanding of the foregoing subject matter, many aspects are described in the form of sequences of acts. At least one of these aspects defined by the claims is performed by an electronic hardware component. For example, it will be recognized that the various acts may be performed by a dedicated circuit or circuitry, program instructions executed by one or more processors, or a combination of both. The description of any sequence of acts herein is not intended to imply that a particular order must be followed to perform that sequence. All methods described herein may be performed in any suitable order, unless otherwise indicated herein or clearly contradicted by the context.

[0082] Although one or more embodiments have been described by way of example and specific embodiments, it should be understood that one or more implementations are not limited to the disclosed embodiments. On the contrary, it is intended to cover various modifications and similar arrangements that are apparent to those skilled in the art. Accordingly, the scope of the appended claims should be given the broadest interpretation to cover all such modifications and similar arrangements.

Claims

1. A system for an artificial intelligence suite for optimizing industrial plant design, the system comprising: One or more processors; And A non-transitory computer-readable medium storing multiple instructions that, when executed, cause the one or more processors to: Integrate engineering knowledge and three-dimensional routing space processing through an artificial intelligence search engine; Prompt a user to input start and end points corresponding to a specified type of conduit via an interactive three-dimensional model of the space in the industrial plant through a graphical user interface; Generate multiple three-dimensional path options from the start point to the end point corresponding to each conduit in the conduit; Determine an optimal path option from the multiple three-dimensional path options corresponding to each conduit in the conduit based on at least one of at least one design constraint or at least one design goal through an artificial intelligence search engine; Modify the optimal path option based on previous user actions in response to previous optimal path options through a machine learning model; Display the modified path option via the graphical user interface via an interactive three-dimensional model of the space in the industrial plant; And In response to a user action, execute a natural language processor to: i) query engineering information, ii) operate the artificial intelligence search engine to generate design options, iii) deploy the user-selected option, or iv) adjust the modified path option.

2. The system of claim 1, wherein the engineering knowledge and three-dimensional routing space processing include at least one of practical rules, requirement considerations, design goals, design constraints, cost metrics, or sustainability metrics.

3. The system of claim 1, wherein the conduit includes a suitable type of channel for conveying at least one of water, liquid, gas, electricity, heat, or air conditioning.

4. The system of claim 1, wherein generating path options for each conduit includes organizing the path options into compatible groups of path options, where each group is ranked according to at least one of the at least one design constraint or the at least one design goal.

5. The system of claim 1, wherein at least one of the at least one design constraint or the at least one design goal of the conduit is input via at least one of a graphical user interface or a natural language processor, and the optimal path option is at least one of constructible, sustainable based on layout options, or inherently conflict-free based on sorting the processing of the conduit.

6. The system of claim 1, wherein the machine learning model performs at least one of the following: learning to guide the artificial intelligence search process from a reference dataset, learning to adjust the modified path option from a natural language processor, or generating additional relevant paths via at least one of a reliability check, a constructibility check, or a feasibility check.

7. The system of claim 1, wherein adjusting via the natural language processor includes adjusting at least one of the at least one design constraint or the at least one design goal, and the natural language processor includes a bidirectional large language model that searches a three-dimensional model associated with engineering information and performs discipline-specific tasks.

8. A computer-implemented method for visualizing data provided by an external source, the computer-implemented method comprising: Integrating engineering knowledge and three-dimensional wiring space processing through an artificial intelligence search engine; Prompting, via a graphical user interface, a user to input a starting point and an ending point corresponding to a specified type of conduit via an interactive three-dimensional space model in an industrial plant; Generating a plurality of three-dimensional path options from the starting point to the ending point corresponding to each of the conduits; Determining, through an artificial intelligence search engine, an optimal path option from the plurality of three-dimensional path options corresponding to each of the conduits based on at least one of at least one design constraint or at least one design goal; Modifying the optimal path option based on previous user actions in response to previous optimal path options through a machine learning model; Displaying the modified path option via the graphical user interface via an interactive three-dimensional model of the space in the industrial plant; And In response to a user action, executing a natural language processor to: i) query engineering information, ii) operate the artificial intelligence search engine to generate design options, iii) deploy an option selected by the user, or iv) adjust the modified path option.

9. The computer-implemented method according to claim 8, wherein the engineering knowledge and three-dimensional wiring space processing include at least one of practical rules, requirement considerations, design goals, design constraints, cost metrics, or sustainability metrics.

10. The computer-implemented method according to claim 8, wherein the conduit includes an appropriate type of channel for conveying at least one of water, liquid, gas, electricity, heat, or air conditioning.

11. The computer-implemented method according to claim 8, wherein generating path options for each conduit includes organizing the path options into compatible groups of path options, wherein each group is ranked according to at least one of the at least one design constraint or the at least one design goal.

12. The computer-implemented method according to claim 8, wherein at least one of the at least one design constraint or the at least one design goal is input via at least one of the graphical user interface or the natural language processor, and the optimal path option is at least one of constructible, sustainable based on a layout option, or inherently conflict-free based on sorting the processing of the conduit.

13. The computer-implemented method according to claim 8, wherein the machine learning model performs at least one of the following: learning from a reference data set to guide the artificial intelligence search process, learning from the natural language processor to adjust the modified path option, or generating additional relevant paths via at least one of a reliability check, a constructibility check, or a feasibility check.

14. The computer-implemented method according to claim 8, wherein adjusting via the natural language processor includes adjusting at least one of the at least one design constraint or the at least one design goal, and the natural language processor includes a bidirectional large language model that performs at least one of the following: searching a three-dimensional model associated with engineering information, or performing a discipline-specific task.

15. A computer program product comprising a non-transitory computer-readable medium having computer-readable program code embodied therein to be executed by one or more processors, the program code including instructions for: Integrating engineering knowledge and three-dimensional routing space processing through an artificial intelligence search engine; Prompting a user to input a start point and an end point corresponding to a specified type of conduit via an interactive three-dimensional model of a space in an industrial plant through a graphical user interface; Generating a plurality of three-dimensional path options from the start point to the end point corresponding to each of the conduits; Determining an optimal path option from the plurality of three-dimensional path options corresponding to each of the conduits based on at least one of at least one design constraint or at least one design goal through an artificial intelligence search engine; Modifying the optimal path option based on previous user actions in response to previous optimal path options through a machine learning model; Displaying the modified path option via the graphical user interface via the interactive three-dimensional model of the space in the industrial plant; And In response to a user action, executing a natural language processor to: i) query engineering information, ii) operate the artificial intelligence search engine to generate design options, iii) deploy the user-selected option, or iv) adjust the modified path option.

16. The computer program product according to claim 15, wherein the engineering knowledge and three-dimensional routing space processing include at least one of practical rules, requirement considerations, design goals, design constraints, cost metrics, or sustainability metrics.

17. The computer program product according to claim 15, wherein the conduit includes a suitable type of passage for conveying at least one of water, liquid, gas, electricity, heat, or air conditioning.

18. The computer program product according to claim 15, wherein generating path options for each conduit includes organizing the path options into compatible groups of path options, wherein each group is ranked according to at least one of the at least one design constraint or the at least one design goal, and wherein at least one of the at least one design constraint or the at least one design goal for the conduit is input via at least one of the graphical user interface or the natural language processor, and the optimal path option is at least one of constructible, sustainable based on layout options, or inherently conflict-free based on sorting the processing of the conduit.

19. The computer program product according to claim 15, wherein the machine learning model performs at least one of the following: learning to guide the artificial intelligence search process from a reference data set, learning to adjust the modified path option from the natural language processor, or generating additional relevant paths via at least one of a reliability check, a constructibility check, or a feasibility check.

20. The computer program product according to claim 15, wherein the adjustment via the natural language processor includes adjusting at least one of the at least one design constraint or the at least one design goal, and the natural language processor includes a bidirectional large language model that performs at least one of the following: searching for a three-dimensional model associated with engineering information, or performing a discipline-specific task.

Images