Systems and methods for intelligent construction management

Abstract

In various examples, the subject matter of this disclosure relates to systems and methods for intelligent construction management. An example method includes: providing access to a construction management platform; ingesting documents related to a construction project; generating a knowledge graph based on the ingested documents, the knowledge graph defining relationships among entities associated with the construction project; obtaining information related to at least one external factor including one or more of weather, news events, or market conditions; determining, using the knowledge graph, a risk that the at least one external factor will cause a delay to a schedule for the construction project; and taking, by at least one automated agent, a corrective action in response to the determined risk.

Description

Docket No. KRN-001WOSYSTEMS AND METHODS FOR INTELLIGENT CONSTRUCTION MANAGEMENTCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 713,262, filed on October 29, 2024, the entire contents of which are incorporated by reference herein.BACKGROUND

[0002] In construction projects, multiple human and organizational stakeholders need to collaborate closely to orchestrate the supply chain, schedule, resources, and relevant prerequisites. They also need to exchange actionable information timely to maximize project speed, productivity, and safety while minimizing cost, waste, and risks.

[0003] Existing construction management tools have many drawbacks. In general, such tools: have a steep learning curve and are inconvenient to use for typical construction workers, foremen, suppliers, and contractors; are incompatible with or missing data from different solution vendors; are incapable of effectively using unstructured construction documents like images, charts, schematics, and tables; are unable to feed real-time information from stakeholders working in the field; and are unable to leverage relevant information from external factors that can cause delays and / or increase costs.

[0004] There is a need for improved systems and methods for managing construction projects.SUMMARY

[0005] In one aspect, the subject matter of this disclosure relates to a computer- implemented method. The method includes: providing access to a construction management platform; ingesting documents related to a construction project into the construction management platform; generating, using the construction management platform, a knowledge graph based on the ingested documents, the knowledge graph defining relationships among entities associated with the construction project; obtaining, using the construction management platform, information related to at least one external factor including one or more of weather, news events, or market conditions; determining, using the knowledge graph, a risk that the at least one external factor will cause a delay to a schedule for the construction project; and taking, by at least one automated agent associated with the construction management platform, a corrective action in response to the determined risk, the corrective1IPTS / 200149918Docket No. KRN-001WO action including at least one of sending a message to a stakeholder for the construction project, modifying an order for a material or service associated with the construction project, or revising the schedule for the construction project.

[0006] In certain examples, the documents can include at least one of a material list, a schedule, a submittal, a drawing, a pre-task plan, an order, a schematic, a request for information, a table, a bill of materials, a job hazard description, or any combination thereof. Ingesting the documents can include: partitioning the documents into regions; extracting metadata from the documents; performing an automated visual scan of the documents; and extracting text from the documents. Generating the knowledge graph can include: providing data extracted from the documents to a predictive model that generates contextualized, vector representations of the data; and constructing the knowledge graph from the contextualized, vector representations. The knowledge graph can be constructed using a material ontology that defines relationships between construction materials and at least one of a construction activity, a construction location, a materials supplier, a materials specification, or a submittal.

[0007] In some implementations, the entities can include at least one of materials, orders, schedules, activities, or prerequisites. Taking the corrective action can include using a voice agent to make an autonomous voice call. The method can include using a graphical user interface in the construction management platform to build a workflow for the construction project, the workflow defining actions to be performed by the at least one automated agent. The actions can include at least one of checking an inventory of a construction material, placing or modifying an order for the construction material, preparing a purchase order, making a phone call to a supplier, or sending an email or text message to a supplier.

[0008] In another aspect, the subject matter of this disclosure relates to a system. The system includes one or more computer processors programmed to perform operations including: providing access to a construction management platform; ingesting documents related to a construction project into the construction management platform; generating, using the construction management platform, a knowledge graph based on the ingested documents, the knowledge graph defining relationships among entities associated with the construction project; obtaining, using the construction management platform, information related to at least one external factor including one or more of weather, news events, or market conditions; determining, using the knowledge graph, a risk that the at least one external factor will cause a delay to a schedule for the construction project; and taking, by at least one automated agent associated with the construction management platform, a corrective action in response to the2IPTS / 200149918Docket No. KRN-001WO determined risk, the corrective action including at least one of sending a message to a stakeholder for the construction project, modifying an order for a material or service associated with the construction project, or revising the schedule for the construction project.

[0009] In some examples, the documents can include at least one of a material list, a schedule, a submittal, a drawing, a pre-task plan, an order, a schematic, a request for information, a table, a bill of materials, a job hazard description, or any combination thereof. Ingesting the documents can include: partitioning the documents into regions; extracting metadata from the documents; performing an automated visual scan of the documents; and extracting text from the documents. Generating the knowledge graph can include: providing data extracted from the documents to a predictive model that generates contextualized, vector representations of the data; and constructing the knowledge graph from the contextualized, vector representations. The knowledge graph can be constructed using a material ontology that defines relationships between construction materials and at least one of a construction activity, a construction location, a materials supplier, a materials specification, or a submittal.

[0010] In various instances, the entities can include at least one of materials, orders, schedules, activities, or prerequisites. Taking the corrective action can include using a voice agent to make an autonomous voice call. The operations can include using a graphical user interface in the construction management platform to build a workflow for the construction project, the workflow defining actions to be performed by the at least one automated agent. The actions can include at least one of checking an inventory of a construction material, placing or modifying an order for the construction material, preparing a purchase order, making a phone call to a supplier, or sending an email or text message to a supplier.

[0011] In another aspect, the subject matter of this disclosure relates to a non-transitory computer-readable medium having instructions stored thereon that, when executed by one or more computer processors, cause the one or more computer processors to perform operations including: providing access to a construction management platform; ingesting documents related to a construction project into the construction management platform; generating, using the construction management platform, a knowledge graph based on the ingested documents, the knowledge graph defining relationships among entities associated with the construction project; obtaining, using the construction management platform, information related to at least one external factor including one or more of weather, news events, or market conditions; determining, using the knowledge graph, a risk that the at least one external factor will cause a delay to a schedule for the construction project; and taking, by at least one automated agent3IPTS / 200149918Docket No. KRN-001WO associated with the construction management platform, a corrective action in response to the determined risk, the corrective action including at least one of sending a message to a stakeholder for the construction project, modifying an order for a material or service associated with the construction project, or revising the schedule for the construction project.

[0012] These and other objects, along with advantages and features of embodiments of the present invention herein disclosed, will become more apparent through reference to the following description, the figures, and the claims. Furthermore, it is to be understood that the features of the various embodiments described herein are not mutually exclusive and can exist in various combinations and permutations.BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In the drawings, like reference characters generally refer to the same parts throughout the different views. Also, the drawings are not necessarily to scale, emphasis instead generally being placed upon illustrating the principles of the invention. In the following description, various embodiments of the present invention are described with reference to the following drawings.

[0014] FIG. 1 is a schematic diagram of a construction management system for an ecosystem of stakeholders, in accordance with certain examples.

[0015] FIG. 2 is a schematic diagram showing module interactions, data flow between components, and user interaction points for a construction management system, in accordance with certain examples.

[0016] FIG. 3 is a schematic diagram of a project info digest module that provides a document processing pipeline, in accordance with certain examples.

[0017] FIG. 4 is a schematic diagram of a hierarchy for a material-specific ontology, in accordance with certain examples.

[0018] FIG. 5 is a schematic diagram of a system for processing documents to prepare a knowledge graph, in accordance with certain examples.

[0019] FIG. 6 is a schematic diagram of an architecture for a workflow builder, in accordance with certain examples.

[0020] FIG. 7 is a schematic diagram of a material order workflow, in accordance with certain examples.

[0021] FIG. 8 is a schematic operational workflow diagram for a construction management system, in accordance with certain examples.4IPTS / 200149918Docket No. KRN-001WO

[0022] FIG. 9 is a schematic diagram of a recommendation mechanism, in accordance with certain examples.

[0023] FIG. 10 is a schematic diagram of a self-learning mechanism, in accordance with certain examples.

[0024] FIG. 11 is a block diagram of an example computer system, in accordance with certain examples.

[0025] While the present disclosure is subject to various modifications and alternative forms, specific embodiments thereof have been shown by way of example in the drawings and will herein be described in detail. The present disclosure should not be understood to be limited to the particular forms disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present disclosure.DETAILED DESCRIPTION

[0026] A description of example embodiments follows. It is contemplated that apparatus, systems, methods, and processes of the claimed invention encompass variations and adaptations developed using information from the embodiments described herein. Adaptation and / or modification of the apparatus, systems, methods, and processes described herein may be performed by those of ordinary skill in the relevant art.

[0027] It should be understood that the order of steps or order for performing certain actions is immaterial so long as the invention remains operable. Moreover, two or more steps or actions may be conducted simultaneously.

[0028] Various aspects and embodiments of a digital construction project orchestration system (referred to as the “Intelligent Construction Digital Manager,” “ICDM,” or “platform”), drives the creation, collaboration, simulation, prediction, execution, and continuous improvement of construction projects. The platform creates and hosts one or more digital networks of agents, including stakeholders (humans, organizations) and virtual agents, acting as proxies for human agents that are linked to construction projects. One example of such a network is a supply chain network of construction that involves general contractors, subcontractors, and suppliers. The agents in the system can dynamically create, update, and monitor project prerequisites (e.g., submittals, material list, POs, and RFIs) and resources, including labor, equipment and space.5IPTS / 200149918Docket No. KRN-001WO

[0029] Aspects of the disclosed subject matter include methods to retrieve, understand and internalize data or files from different solution vendors to be usable by the platform and agents, as well as to extract and index structured information from charts, images, schematics and tables. Once collected and stored, this information is accessible to query agents as well as for generating analytics. Project prerequisites can be identified and linked to each other across projects and stakeholders to reduce the time needed for project planning, scheduling and arranging resources. Multimodal natural interfaces can be created that allow stakeholders to view the precise status of a project and make changes to any project artifacts using voice, text, and / or email from anywhere, anytime, on any device.

[0030] The system continuously receives, aggregates and indexes textual data feed (e.g., news), time series data (e.g., signals from loT sensors), and / or visual data (e.g., cameras, etc.) to generate relevant analytics and forward to the relevant system agents. The system can then direct a network of agents (human and / or virtual) to execute actions at the right time to provide optimal throughput of projects. This results in a digital representation of a construction project (e.g., a digital twin) that can run close to real-life simulation and / or replay of the construction process for prediction and analysis purposes. In some examples, the system can receive, digest and analyze relevant internal and external data (both structured and unstructured) in real time and generate actionable information that is sent to the relevant stakeholder(s).

[0031] In certain examples, the disclosed system includes an intelligent voice agent specifically designed for supply chain management. The voice agent is capable of conducting autonomous voice calls to suppliers, subcontractors, and port authorities to gather real-time information on material availability, shipment delays, and logistics updates. The voice agent can automatically update procurement logs and analytics dashboards with gathered information.

[0032] Additionally or optionally, a material-specific ontology can be implemented (e.g., within a knowledge graph) to extract and organize material data from construction documents (e.g., drawings, schedules, specifications). The ontology can filter out non-material information to create a rich, focused knowledge base that reduces noise and improves material tracking accuracy.

[0033] Additionally or optionally, a no-code agentic workflow builder can enable users to create custom supply chain management and analytics workflows without programming knowledge. The workflow builder can allow construction managers or other users to6IPTS / 200149918Docket No. KRN-001WO configure agent behaviors, data collection sequences, and automated decision-making processes through an intuitive visual interface.

[0034] In various examples, an “agent” (alternatively referred to herein as an “automated agent,” an “assist agent,” or a “virtual agent”) can be or include a software program and / or a computer-implemented process that is configured to perform tasks autonomously on behalf of a user or system. The agent can utilize machine learning or artificial intelligence and / or can interact with humans or other entities through chat or voice.System Overview and Architecture

[0035] FIG. 1 is a schematic diagram of a construction management system 100 for an ecosystem of stakeholders, in accordance with certain examples. The system 100 includes a construction project info digest (PID) module 110, a time-series event digest (TSED) module 112, a multi-tenant knowledge database 114, a processing / agent layer 116, an agent coordinator 118, a construction project digital twin 120, and an operation control center 122. Stakeholders for the ecosystem can include one or more of suppliers 124 (e.g., companies that provide construction materials), logistics providers 126 (e.g., truck drivers, shipping companies, warehouses, etc.), sub-contractors 128, and general contractors 130. External data sources, including a 3rdparty solution source selector 132 and a real-time data source selector 134 feed data into the system 100, which processes the data and coordinates with various construction stakeholders, as described herein.

[0036] FIG. 2 is a detailed system architecture 200 showing module interactions, data flow between components, and user interaction points for a construction management system (e.g., the system 100), in accordance with certain examples. The architecture 200 includes the PID module 110, the TSED module 112, a multi-modal interface switch (MMIS) 214, a multitenant knowledge base (MTKB) 216, one or more construction agents 218 (e.g., an automated or Al -enabled agent), a construction engine 220 (e.g., including the operation control center 122), and a smart link 222 (e.g., including a knowledge graph and / or a predictive model). One or more external data sources 224 (e.g., including third party solution sources and / or real-time data) can be provided to system components through an access handler 226 (e.g., managing or authenticating log in credentials). One or more users 228 (e.g., subcontractors 128 or general contractor 130) can access the system through an access handler 230. In various examples, users 228 of the system can include a tenant or group of users who are working together on a construction project. The PID module 110 and / or other7IPTS / 200149918Docket No. KRN-001WO modules or components of the architecture 200 can be implemented with software, firmware, and / or one or more processors.PIP Module

[0037] FIG. 3 is a schematic diagram of the PID module 110 providing a document processing pipeline, in accordance with certain examples. Documents and / or other data from external data sources 224 (e.g., project software, a project info database, and / or 3rd-party systems) flows through the access handler 226 into the PID module 210. The PID module 210 performs partition and metadata extraction 318 on the documents (e.g., to divide the documents into sections or partitions and obtain metadata), and then processes the documents or portions thereof using a visual algorithm 320 (e.g., that visually scans the documents for relevant construction project information) and textual extraction 322 (e.g., to extract text from the documents). The visual algorithm 320 and the textual extraction 322 can occur on parallel processing paths. Output from the visual algorithm 320 and the textual extraction 322 is fed into a predictive model 324 (e.g., a large language model or LLM), which generates outputs stored in a vector store 326 and a knowledge graph 328. The vector store 326 and the knowledge graph 328 are then accessible to artificial agents 330 (e.g., including one or more construction agents 218) for use in generating, accessing, or using the construction project digital twin 120.

[0038] A typical construction project (e.g., a commercial, industrial, or infrastructure construction project) involves sharing massive amounts of information between stakeholders (e.g., general contractors, subcontractors, suppliers, safety supervisors, project engineers, etc.), across a variety of documents types (e.g., including or relating to material lists, schedules, submittals, drawings, pre-task plans, orders, schematics, requests for information, tables, bills of materials, job hazards, etc.). These documents can be hosted, managed, and / or provided by solutions vendors, which can be or include, for example, companies, individuals, or other entities that provide services or materials related to managing or executing construction projects or aspects thereof. A solution vendor can be or include, for example, an entity that provides software or services related to construction project management, order management, or financial management.

[0039] In various examples, the PID module 110 is configured to extract information from a variety of document types (e.g., provided by external data sources 224). Such document types can include, for example, PDF files, Word files, Excel files, rich text files, and other file types or file formats. Information from the documents can be contextualized and represented8IPTS / 200149918Docket No. KRN-001WO as mathematical vectors in high dimensional space (e.g., using the model 324). The vectors can be stored in the vector store 326. The knowledge graph 328 can be built to represent or map relationships between documents and entities (e.g., stakeholders, vendors, etc.), materials, activities, or other items associated with the construction project.

[0040] In some instances, the PID module 110 can preserve data privacy of each data provider and application user. One or more users (e.g., in a tenant of users for the project) can set a desired access and / or permission control for each data source. The PID module 110 can be configured to access only data that are granted read permission.

[0041] Additionally or alternatively, the PID module 110 can identify each document type and, before information is extracted, the PID module 110 can partition the document into regions (e.g., by performing partition and metadata extraction 318), with each region corresponding to, for example, a portion of text, a table, a diagram, an image, etc. The coordinates of each region can be recorded and / or embedded as metadata associated with information extracted for the region.

[0042] In various examples, the PID module 110 can use one or more algorithms to extract information from documents. An algorithmic selection method can be used to determine the best algorithm to use, depending on a complexity of structures within the target object or document region. For example, an object detection (OD) algorithm can detect elements within each document region or object (e.g., table rows and / or columns). Those elements can be passed to Optical Character Recognition (OCR) to convert text in bitmap format into text representations understandable by machines. The elements or associated information can then be represented with structural data. Multimodal vision LLM models can be used to represent tables or other objects that are less structured.

[0043] In some examples, the PID module 110 (e.g., using the model 324) can generate contextualized vector representations or embeddings for objects or chunks of extracted information. The embeddings can be or include mathematical representations of real world objects (e.g., text, images, voice, etc.). The embeddings can be stored in the vector store 326 to be retrieved using algorithms (e.g., proximity search) during a retrieval process.

[0044] In certain implementations, the knowledge graph 328 can be generated and used for organizing information about entities and their relationships. The PID module 110 can apply or use the knowledge graph 328 for requests for information (RFIs) or submittal documents that are versioned and linked (e.g., as supplementary documents or updates). Knowledge graph 328 entities for documents can be created and linked to represent those versions and9IPTS / 200149918Docket No. KRN-001WO inter-document relationships. Additionally or alternatively, the PID module 110 can apply or use the knowledge graph 328 for project entities (e.g., materials, orders, schedules, activities, and / or prerequisites) that can be represented and linked in the knowledge graph 328 by nodes, edges, and labels. The knowledge graph 328 can allow one or more agents (e.g., retrieval-augmented generation or RAG agents) to be able to efficiently retrieve project entities based on a current usage context of a user prompt. For example, when a user is focusing on a specific project ID (e.g., an identifier for a material, a vendor, or an activity for a construction project), highly related entities represented by the knowledge graph 328 (e.g., entities that are directly connected or in close proximity) can be ranked higher in a search result.

[0045] In various examples, the PID module 110 preserves access control policies (e.g., using role-based access control or RBAC) of the source data (e.g., documents or data extracted from documents). In a typical construction project, different types of documents (e.g., drawings, RFIs, and submittals) can have different permissions granted to various stakeholders. For example, subcontractors may be unable to open and / or read biddings but may be able to open and / or read drawings. The embeddings generated for chunks of extracted information, as described herein, can retain the permission attributes of the corresponding document type. To make sure there are no data leaks, the embeddings of different document types can be in different vector spaces and / or stored in different vector tables (e.g., in database 114). In the retrieval step, tables accessible by the user can be merged before the vector space search. This guarantees no unauthorized data is viewable by the user.Time Series Event Digest (TSED)

[0046] Referring again to FIG. 2, in various examples, the TSED module aggregates project relevant real-time information of factors that have direct impacts on cost, time, and / or success of the project. Information processed by the TSED module can include information related to external factors that may or may not be controllable by individuals associated with the construction project (e.g., the general contractor 130). Such information can include, for example, information related to weather, news events, supply markets (e.g., raw materials), labor markets, logistics (e.g., transportation and / or storage), demand markets (e.g., real estate), indirect events (e.g., political, geopolitical, public health, economics, etc.), direct project events (e.g., order shipments, activities updates, etc.), Internet of things (loT) sensor events (e.g., video camera, RFID, etc.), or any combination thereof.10IPTS / 200149918Docket No. KRN-001WO

[0047] Based on the needs of agents in each project, an operation control center (OCC) (e.g., the operation control center 122) can configure “windows” of TSED and / or can consider a geographical location of data. The OCC can define or determine a maximum lookahead time (TLA), a maximum lookback time (TLB), and / or a temporal granularity of data for each data source. The data may be stored in a NoSQL database (or other database) and indexed by time and geographical location.

[0048] Different agents (e.g., virtual agents, automated computer-implemented processes, and / or agents that utilize machine learning or artificial intelligence) associated with the project can pull required data needed for predictive analysis. Lor example, a cost analysis agent can pull (i) price of materials data from corresponding markets and / or (ii) regional labor market data to forecast the cost of materials. Additionally or alternatively, scheduling risk agents can retrieve logistics data (e.g., shipping time and / or weather forecast data) to predict a risk of delays to the project schedule. Such data can be retrieved and analyzed (e.g., by the cost analysis agent and / or scheduling risk agents) daily or at other time intervals. The cost analysis agent, scheduling risk agent, and / or other agents described herein can utilize one or more machine learning models and / or artificial intelligence to make predictions and / or take action.Multi-tenant Knowledge Base (MTKB)

[0049] In various examples, the multi-tenant knowledge base (MTKB) can provide each tenant (e.g., a group of users for a construction project) with exclusive rights to structured data, unstructured data, and tenant specific models. Access to databases used by the systems and methods (e.g., the database 114, including vector databases, relational databases, and / or time-series databases) can be limited (e.g., using tables or constraints) based on the rights or permissions of the agent or associated tenant. Such agents may be or include virtual agents (e.g., deployed with agentic Al), each inheriting the access rights of the human agent that it acts on behalf of.Multi-Modal Interface Switch (MIS)

[0050] In certain examples, the MMIS module 214 functions as a message exchange (e.g., operating 24 hours / day, 7 days / week) that connects to multiple interface modalities with humans. Modalities include, but are not limited to, text messaging, voice messaging, email, chatbot, chat application interfaces, or any combination thereof. In construction, different roles (e.g., project managers, supervisors, drivers, and field workers) can have different working conditions and constraints. The MMIS module is configured to select a most optimal11IPTS / 200149918Docket No. KRN-001WO interface modality based on the role and / or preferences of the user. For example, a subcontractor working in the field may prefer to receive and / or provide messages only by phone or text messaging. The MMIS module can leverage one or more predictive models (e.g., a large language model or LLM) specifically fine-tuned for each modality and / or can use a prompt and response format that is most appropriate for the modality.Construction Agent Team

[0051] In various examples, a set of construction agents 218 work independently and / or collaboratively to execute a variety of functions associated with a construction project. The agents can integrate with and direct actions in one or more construction software tools (e.g., using function calling methods). The construction software tools can include, for example, material tagging tools, material ordering tools, supplier notification tools, software provided by third party solution vendors, or any combination thereof. Additionally or optionally, the agents can plan a sequence of steps to execute operations associated with the construction project. Such operations can include, for example, ordering materials for one or more scheduled activities, changing an activity schedule and / or related prerequisites, etc. Additionally or optionally, the agents can collaborate with one or more other agents to perform complex operations in parallel. For example, material ordering agents can collaborate with scheduling agents and human agents (e.g., suppliers and / or subcontractors) to execute or implement supply chain network changes (e.g., as ordered by the general contractor 130).

[0052] In certain instances, an agent team 218 can interact with a user through multiple modalities including, for example, short messaging (SMS), voice messaging (VMS), email, and / or chat applications. The agents can support a most natural or preferred method of interaction with each respective user (which may be different from user to user), such that the method can require minimal learning or effort on the part of the user. For example, SMS usually are dominated by short form messages and acronyms, whereas emails can be dominated by long sentences, and voice messages can be a mix of dialects and noises. A user query dialect adapter agent can be trained to transform and enhance raw information to / from the user. Based on the source modality, the adapter agent can fill in missing information (e.g., convert an acronym to full form), remove redundant words, and / or perform denoising or error correction, such that a single nominal instruct model can determine the intent of the user before performing the functions specified herein.12IPTS / 200149918Docket No. KRN-001WOConstruction Engine (CE)

[0053] Referring to FIGS. 1 and 2, in certain examples, the construction engine 220 can generate and / or utilize the construction project digital twin 120, which can be a digital carbon copy of a construction project. The construction engine 220 can ensure that each and every agent, schedule activity, prerequisite, material, projected start date, end date, lead time, etc. can be represented in the digital twin 120. External factors leading to delay and / or a cost increase can be added to the digital twin 120 during project simulation or re-run. When a new project is created, all entities and dependencies can be initialized, including schedule activities, start date, sub-contractors, suppliers, materials, etc.

[0054] With previous approaches, a project schedule timeline and cost analysis can be performed before the construction starts; however, there are some limitations to the preconstruction timeline and cost analysis. For example, certain risk factors can disproportionately affect the full project schedule (e.g., because of an unclear dependency chain). It can be difficult with previous approaches for a planner or project manager (e.g., the general contractor 103) to identify how risk factors may affect specific parts or phases of the project, so that corresponding measures can be taken. For example, with previous approaches, a shortage of drivers can affect the delivery of cement or other materials, which can postpone downstream activities, like plumbing and electrical work. When construction is underway, it may be unclear how certain delays will impact downstream tasks. Further, with previous approaches, a project manager may lack tools to reliably realign or reprioritize activities to maximize project efficiency. Additionally, after a construction project is completed, there is generally no good way to perform a post-mortem so that staff can be trained to better identify critical weaknesses in the schedule for future projects.

[0055] Advantageously, the these challenges with previous approaches can be overcome by the systems and methods described herein. The systems and methods (e.g., utilizing the construction engine 220 and the digital twin 120) can provide full replication of all entities and schedules and can reuse all prerequisites (e.g., documents, submittals, RFIs, drawings) of a project. The systems and methods can allow external environmental factor simulators or generators to adjust project variables (e.g., using variability knobs or controls), such as start time, end time, lead time, material quantities, etc. The systems and methods can utilize instrumentation to measure and record changes in value of each variable when other conditions change. An external environmental factor generator can be used to adjust the variable knobs of the project. For example, a global logistic simulator can be used to adjust13IPTS / 200149918Docket No. KRN-001WO the material supply lead time according to the forecast of transportation time of materials after submitting an order. This allows the systems and methods to automatically adapt and adjust schedules to compensate for delayed shipments, bad weather, or other factors that could otherwise increase overall project costs or construction times.Operation Control Center (OCC)

[0056] Referring to FIG. 1, in various examples, the operation control center 122 can operate as a control plane that allows users to create and manage projects and / or control access to the system. The operation control center 122 can provide a dashboard that allows users to view the status and timeline of schedules. The operation control center 122 can orchestrate and manage a virtual agent team (e.g., using the agent coordinator 118). The operation control center 122 can control the permission of access to data, either by human or virtual agents.Material Ontology

[0057] In various examples, the system described herein (e.g., system 100) implements and uses a specialized, material-focused ontology designed specifically for construction document ingestion and knowledge graph construction. The ontology provides a structured framework for extracting, classifying, and relating material information (e.g., related to construction materials) from diverse construction documents, including construction drawings, project schedules, and specification documents. The ontology can be implemented and / or executed using the PID module 110 or other system component (e.g., the operation control center 122).

[0058] In certain implementations, the material ontology can define hierarchical relationships and properties specific to construction materials. For example, construction materials can be classified into categories (e.g., structural, mechanical, electrical, plumbing, and / or finishes) with properties including, for example, material type, manufacturer, model number, specifications, quantities, unit costs, lead times, installation requirements, and / or dependencies. Additionally or optionally, the ontology can include or utilize document-type - specific extraction patterns. For example, from construction drawings, information can be extracted according to material callouts, detail references, legend symbols, dimension specifications, and / or material specifications (e.g., referenced in title blocks). From schedules, information can be extracted according to, for example, door schedules, window schedules, finish schedules, and / or equipment schedules (e.g., with associated material14IPTS / 200149918Docket No. KRN-001WO specifications). From specification documents (e.g., CSI MASTERFORMAT sections), information can be extracted according to detailed material properties, performance requirements, manufacturer requirements, installation procedures, and / or quality standards.

[0059] In some examples, the ontology can define relationships between materials and other project entities (e.g., orders, schedules, suppliers, submittals, specifications, activities, location, and / or prerequisites). Materials can be related to activities, for example, by identifying materials that are desired or required for specific construction activities. Materials can be related to location, for example, by determining where materials are installed (e.g., floor, zone, or room). Materials can be related to supplier, for example, by identifying approved suppliers and manufacturers for each material. Materials can be related to submittals, for example, by linking materials to their submittal documents and / or approval status. Materials can be related to specifications, for example, by mapping materials to relevant specification sections in documents or to specifications that define requirements for construction activities.

[0060] In some implementations, the ontology can perform noise filtering and scoping by employing intelligent filtering mechanisms to scope extraction only to material-relevant data. For example, contextual filtering can be performed using natural language processing to identify material-relevant sections within documents and optionally filter out or remove nonmaterial content such as general notes, code requirements, and procedural instructions that do not directly describe materials. Semantic disambiguation can be performed by distinguishing between material mentions (e.g., “concrete” as a construction material) versus non-material uses of the same terms (e.g., “concrete plans” meaning definitive plans). Attribute extraction can be performed by extracting material-specific attributes while ignoring irrelevant document content, for example, by focusing on quantities, dimensions, colors, finishes, grades, strengths, certifications, and / or performance specifications. Hierarchical scoping can be performed by maintaining parent-child relationships (e.g., “steel beam” is a child of “structural steel” which is a child of “metals”) to enable both granular and high-level material queries.

[0061] The ontology can be used by the PID module 110 to construct the knowledge graph 328. The process can begin with document ingestion, where construction documents are processed through the partitioning and extraction pipeline of the PID module 110. Ontology- guided extraction can be performed wherein material ontology guides the extraction process, identifying material -re levant content and ignoring non-material sections. Entity recognition15IPTS / 200149918Docket No. KRN-001WO can be performed to identify and classify material entities according to the ontology’s taxonomy. Relationship extraction can be performed to identify and map relationships between materials and other project entities. Extracted entities and relationships can be stored as nodes and edges in the knowledge graph 328. This focused, material-specific approach provides the knowledge graph 328 with higher precision and lower noise compared to generic extraction methods. The material-specific ontology described herein enables more accurate material tracking, better procurement analytics, and improved supply chain management by ensuring the knowledge graph 328 contains rich, focused material data rather than being diluted with irrelevant information from construction documents.

[0062] FIG. 4 is a schematic diagram of a hierarchy 400 for a material-specific ontology 410, in accordance with certain examples. The hierarchy 400 includes taxonomic classification of construction materials into five major categories 412 and representative subcategories 414. The major categories 412 in this example are structural, mechanical, electrical, plumbing, and finishes. The hierarchy 400 enables both granular material tracking and high-level category analysis while maintaining parent-child relationships for efficient knowledge graph queries.

[0063] FIG. 5 is a schematic diagram of a system 500 for processing documents to prepare a knowledge graph (e.g., the knowledge graph 328), in accordance with certain examples. The system 500 includes a processing pipeline 510 (e.g., in the PID module 110) that processes construction documents 512, which can include drawings, schedules, specifications, or other information (e.g., provided by external data sources 224). The processing pipeline 510 can perform document partitioning, classification and / or extraction on the documents 512, using an ontology 514 (e.g., the ontology 410 and / or including material taxonomy, extraction rules, and / or noise filters) for guidance. Information extracted by the pipeline 510 can be provided to an entity processor 516 to perform entity recognition and attribute extraction (e.g., using the visual algorithm 320 and the textual extraction 322). The entity information and attributes can be include in a knowledge graph 518 as nodes and relationship edges.Agentic Workflow Builder

[0064] In various examples, the construction management systems and methods can include or utilize an intuitive, no-code workflow builder (e.g., as part of the PID module 110) that enables users to create, configure, and deploy custom agentic workflows for supply chain management and analytics without requiring programming knowledge. The workflow builder16IPTS / 200149918Docket No. KRN-001WO can include or utilize a visual design interface (e.g., a graphical user interface) that democratizes access to advanced agent automation capabilities for construction professionals.

[0065] FIG. 6 is a schematic diagram of an architecture 600 for the workflow builder, in accordance with certain examples. A user interface (UI) layer 610 (e.g., atop layer) can provide a visual canvas and node palette. A node types available layer 612 (e.g., a middle layer) can provide available workflow components, including, for example, trigger, data collection, processing, decision, action, and agent invocation. A workflow engine layer 614 (e.g., a bottom layer) can process a configured workflow using a parser, compiler, and executor. A system integration layer 616 can connect the architecture to all platform components, including, for example, a voice agent, a project schedule, an analytics dashboard, and / or a knowledge graph.

[0066] FIG. 7 is a schematic diagram of a material order workflow 700, in accordance with certain examples. The workflow 700 can achieve trigger-based automation that checks inventory levels, invokes voice agents to contact suppliers (e.g., by phone, email, or text messaging), processes responses (e.g., from suppliers), makes decisions based on quotes and criticality, generates purchase orders, and alerts stakeholders as needed. In the depicted workflow 700, one or more actions of automated agents can be triggered at certain periods of time (e.g., a certain time of day) or at certain stages of a project. For example, at 8:00 am each day, an agent can check inventory levels (step 710). If inventory is low, an agent can contact (step 712) a supplier (e.g., by phone, text, or email) to obtain a quote for additional materials. After determining (step 714) that the quote is acceptable, an agent can generate a purchase order (step 716) and / or notify (step 718) a project manager (e.g., the general contractor 130). The workflow 700 can be created visually without any programming, using the no-code workflow builder described herein. For example, a user of the system can construct the workflow 700 in a graphical user interface by (i) selecting action items, decision points, and / or other workflow elements and (ii) adding arrows that connect the elements, as depicted.

[0067] FIG. 8 is a schematic operational workflow diagram 800 for the construction management systems and methods described herein and depicts end-to-end project lifecycle and system interactions, in accordance with certain examples. A construction project can include five sequential phases: project creation 810 (e.g., with input sources from 3rd party sources, such as PROCORE / GOOGLE DRIVE and Control Center Portal via Mobile / Web App), project configuration 812, pre-project planning 814, in-project monitoring 816, and17IPTS / 200149918Docket No. KRN-001WO closure 818. A plurality of virtual agents are available for assisting with project planning 820 and operations / administration 822. The agents can include, for example, a schedule agent (e.g., that monitors, develops, and / or modifies project schedules), a material agent (e.g., that manages construction material quantities and orders), a QA agent (e.g., that provides quality assurance), an order agent (e.g., that places or modifies orders), a safety agent (e.g., that identifies and addresses safety issues), and / or other agents. The system provides SMS 824, email 826, and / or voice 828 interface modalities through which external stakeholders (e.g., drivers, foremen, suppliers, etc.) can interact with the system (e.g., without having to log in). Connecting arrows in the diagram 800 demonstrate how information can flow between project phases, agents, and communication channels, for orchestrating construction project management from inception to completion.Project Creation

[0068] New projects can be created through a variety of workflows. For example, a project can be imported from external project planning software, project documents (e.g., PDF, PROCORE, AUTODESK, EXCEL), or other external sources. When the import is from PDF, import software can identify and extract key information from the PDF file using object detection and optical character recognition techniques to pull textual and numerical information from the project plan. The import software can identify common columns or other features (e.g., schedule activity descriptions, start time, end time, and work breakdown structure) to map to the platform’s canonical project format. The import software can prompt the user when columns cannot be clearly mapped to internal representation. Custom columns can be created in such instances. Additionally or optionally, a new project can be created through a web app or mobile app user interface. In some examples, a digital agent can create a new project upon request by a user. The agent can determine (e.g., from chat or other messages) the user’s intent and can generate new projects that meet the user’s requirements or specifications.Project Configuration

[0069] Once a projected has been created (e.g., to define project schedule activities and high level schedules), more detailed configurations can be added. This can involve linking or tagging specific elements of the project together to establish dependencies. One example is the linking of specific building materials needed for a specific construction activity, such as installation of heat ventilation air conditioning (HVAC) at a specific location linked to specific models of an Air Handling Unit (AHU). A typical project can involve linking18IPTS / 200149918Docket No. KRN-001WO thousands of materials and activities. Such linking can be performed using a smart link (e.g., the smart link 222), a knowledge graph (e.g., the knowledge graph 328), and / or one or more predictive models, as described herein. Advantageously, the systems and methods described herein can utilize material assist agents that automatically recommend and tag materials to selected activities.

[0070] Referring to FIG. 9, in certain examples, a recommendation mechanism 900 used by agents to link materials and activities can be based on a hybrid search process. Starting from a task or activity 910, two parallel processing paths can execute: (1) a dense embedding model 912 that performs a semantic search to find materials with similar meaning and context, and (2) a sparse model 914 that matches exact unique identifiers like SKUs and serial numbers. Both paths feed into a “hybrid search” component 916 that combines the results. The process can be bidirectional, as indicated by the arrow between the hybrid search output and “material,” such that users can select either an activity 910 to find materials 918 or a material 918 to find activities 910. This hybrid approach balances semantic understanding with precise identifier matching for superior recommendation accuracy. This greatly reduces the chance of user mistakes and increases efficiency. Similar methods can be applied to linking other project elements or prerequisites, such as linking submittals to schedules, linking submittals to materials, linking detailed schedules to high level schedules, etc.

[0071] In various examples, the system is self-learning to continuously improve the precision and recall rate of the recommendations for each project. The self-learning capability can be achieved by maintaining a knowledge base (e.g., stored in the database 114) of the construction industry. The knowledge base can include a library of construction articles (e.g., published articles, construction handbooks for electrical systems, air conditioning, etc.) and updated information (e.g., material classifications) from academia or other sources. The library can be expanded to include new information over time. A construction base model can be developed that is trained and / or re-trained based on information in the library.

[0072] Additionally or optionally, the self-learning capability can be achieved or improved by fine-tuning the construction base model for specific projects. Training materials for finetuning the model for a project can include, for example, documents that are specific to the project, such as RFIs, specs, and the like. Such training materials can be proprietary and closely relevant to each project. The documents can be pulled from a database (e.g., the database 114) and / or construction information management systems.19IPTS / 200149918Docket No. KRN-001WO

[0073] Additionally or optionally, the self-learning capabilities can be achieved or improved as the construction base model is being used by users. For example, when a user is using model recommendations and / or linking materials to activities, the system can generate model training data by tracking or monitoring the user’s selections to link materials and activities. The additional training data can be used to fine-tune and / or retrain the model (e.g., weekly or monthly).

[0074] Referring to FIG. 10, in various examples, a self-learning mechanism 1000 for a construction based model 1010 can be achieved by obtaining inputs from three sources: (1) an industry knowledge base 1012 containing published construction handbooks, material classifications, and academic research that is periodically incorporated through training; (2) per-project documents 1014 including proprietary submittals, RFIs, and specifications that fine-tune the model for project-specific terminology and requirements, and (3) user interaction data 1016 capturing actual material -to-activity tagging decisions that provide human-in-the-loop feedback for identifying true / false positives / negatives. Information from all three sources can continuously flow to the model 1010, creating a feedback loop that progressively improves precision and recall rates for recommendations 1018 over time.

[0075] In various examples, different companies and projects have differences in how materials and / or activities are described. Company and project specific models may be developed through the periodic gathering and analysis of project data such as tagged materials and activities for model fine-tuning.

[0076] In certain instances, a configuration step can be performed that links projects to stakeholders such as subcontractors, suppliers, drivers, etc. This can establish a network of human agents that interact with the software. Email addresses and / or phone numbers of stakeholders can be recorded, allowing the system’s multimodal interface switch 214 to send and / or receive messages (e.g., email or voice messages) to and from the stakeholders.Pre-Project Planning

[0077] In various examples, the pre-project planning phase can involve confirming project details (e.g., schedules, materials, equipment, spaces) with each of the stakeholders. After materials are linked to scheduled activities, a user can select suppliers that will provide the materials. At this step, materials-related actions (e.g., create order, release order, complete order, delivery setup, etc.) can be taken and the dates and quantities can be finalized and verified by the system.20IPTS / 200149918Docket No. KRN-001WO

[0078] In some examples, pre-project planning can involve using a safety agent to analyze a pre-task job hazard analysis (JHA) filed by a safety officer. The agent can inform a safety officer and / or project manager if the JHA and pre-task planning are not adequate or up to code.In-Project Monitoring

[0079] Once the project starts, the systems and methods described herein can enter a monitoring mode, where execution of the project is continuously monitored. Reminders, alerts, and / or recommendations can be sent to the stakeholders (e.g., by automated agents) to take action at appropriate times, and responses from the stakeholders can be recorded and forwarded to other stakeholders of the project.

[0080] The system can monitor and / or address several aspects of the construction project during the in-project monitoring phase. For example, the system can provide recommendations to users to ensure certain activities are performed ahead of time (e.g., ordering materials). Additionally or alternatively, the system can perform daily checks of activity-related dependencies completion. If dependencies are not met at a required time, alerts can be sent to the user (e.g., mentioning that materials were not ordered by a target lead time). In some instances, the system can notify stakeholders regarding completion of actions (e.g., delivery of materials). Any of the preceding tasks can be performed using one or more automated agents. The system can perform daily checks of external dependencies (e.g., news, weather, and market) and feed corresponding data to risk and cost analysis agents. The agents can compute a probability of delays and notify stakeholders accordingly.Machine Learning and Neural Networks

[0081] Machine learning is a method of teaching computers to learn and make decisions on their own, without explicitly being programmed to perform a specific task. It involves feeding a large amount of data into a computer program, which then uses statistical analysis to identify patterns and relationships within the data. The goal is to enable the program to make predictions or decisions based on these patterns and relationships, without being explicitly told how to do so.

[0082] Neural networks are a type of machine learning algorithm that are inspired by the structure and function of the human brain. They consist of layers of interconnected “neurons,” sometimes called nodes, which process and transmit information. Each neuron receives input from other neurons, processes it, and passes it on to other neurons in the next layer.21IPTS / 200149918Docket No. KRN-001WO

[0083] The layers in a neural network refer to the layers of interconnected neurons. There are typically multiple layers in a neural network, with the input layer receiving the raw data and the output layer producing the final prediction or decision. Between the input and output layers, there are one or more hidden layers, which process the data and pass it on to the next layer.

[0084] By training a neural network on a large dataset, the connections between neurons (called “weights”) can be adjusted to improve the network’s ability to make predictions or decisions. To train a neural network, the data is fed through the network and the output is compared to the desired result. If the output is not accurate, the weights are adjusted to reduce the error. This process is repeated multiple times, with the network continually adjusting the weights to improve its accuracy. Once the network has been trained, it can be used to make predictions or decisions on new data, based on the patterns and relationships it has learned from the training data.

[0085] In various examples, “machine learning” can refer to the application of certain techniques (e.g., pattern recognition and / or statistical inference techniques) by computer systems to perform specific tasks. Machine learning techniques (automated or otherwise) may be used to build data analytics models based on sample data (e.g., “training data”) and to validate the models using validation data (e.g., “testing data”). The sample and validation data may be organized as sets of records (e.g., “observations” or “data samples”), with each record indicating values of specified data fields (e.g., “independent variables,” “inputs,” “features,” or “predictors”) and corresponding values of other data fields (e.g., “dependent variables,” “outputs,” or “targets”). Machine learning techniques may be used to train models to infer the values of the outputs based on the values of the inputs. When presented with other data (e.g., “inference data”) similar or related to the sample data, such models may accurately infer the unknown values of the targets of the inference data set. Such models can be referred to herein as “machine learning models,” predictive models,” or “computer-implemented models.”

[0086] Features can also have data types. For instance, a feature can have an image data type, a numerical data type, a text data type (e.g., a structured text data type or an unstructured (“free”) text data type), a categorical data type, or any other suitable data type. In the above example, the feature of a shape extracted from an image of a cell can be of an image datatype. In general, a feature’s datatype is categorical if the set of values that can be assigned to the feature is finite.22IPTS / 200149918Docket No. KRN-001WO

[0087] As used herein, the “development” of a machine learning model may refer to construction of the machine learning model. Machine learning models may be constructed by computers using training data sets. Thus, “development” of a machine learning model may include the training of the machine learning model using a training data set. In some cases (generally referred to as “supervised learning”), a training data set used to train a machine learning model can include known outcomes (e.g., labels or target values) for individual data samples in the training data set. For example, when training a supervised computer vision model to detect images of cats, a target value for a data sample in the training data set may indicate whether the data sample includes an image of a cat. In other cases (generally referred to as “unsupervised learning”), a training data set does not include known outcomes for individual data samples in the training data set.

[0088] Following development, a machine learning model may be used to generate inferences with respect to “inference” data sets. For example, following development, a computer vision model may be configured to distinguish data samples including images of cats from data samples that do not include images of cats. As used herein, the “deployment” of a machine learning model may refer to the use of a developed machine learning model to generate inferences about data other than the training data.Computer Implementations

[0089] In some examples, some or all of the processing described above can be carried out on a personal computing device, on one or more centralized computing devices, or via cloudbased processing by one or more servers. Some types of processing can occur on one device and other types of processing can occur on another device. Some or all of the data described above can be stored on a personal computing device, in data storage hosted on one or more centralized computing devices, and / or via cloud-based storage. Some data can be stored in one location and other data can be stored in another location. In some examples, quantum computing can be used and / or functional programming languages can be used. Electrical memory, such as flash-based memory, can be used.

[0090] FIG. 11 is a block diagram of an example computer system 1100 that may be used in implementing the technology described in this document. General-purpose computers, network appliances, mobile devices, or other electronic systems may also include at least portions of the system 1100. The system 1100 includes a processor 1110, a memory 1120, a storage device 1130, and an input / output device 1140. Each of the components 1110, 1120, 1130, and 1140 may be interconnected, for example, using a system bus 1150. The processor23IPTS / 200149918Docket No. KRN-001WO1110 is capable of processing instructions for execution within the system 1100. In some implementations, the processor 1110 is a single-threaded processor. In some implementations, the processor 1110 is a multi -threaded processor. The processor 1110 is capable of processing instructions stored in the memory 1120 or on the storage device 1130.

[0091] The memory 1120 stores information within the system 1100. In some implementations, the memory 1120 is a non-transitory computer-readable medium. In some implementations, the memory 1120 is a volatile memory unit. In some implementations, the memory 1120 is a non-volatile memory unit.

[0092] The storage device 1130 is capable of providing mass storage for the system 1100. In some implementations, the storage device 1130 is a non-transitory computer-readable medium. In various different implementations, the storage device 1130 may include, for example, a hard disk device, an optical disk device, a solid-date drive, a flash drive, or some other large capacity storage device. For example, the storage device may store long-term data (e.g., database data, fde system data, etc.). The input / output device 1140 provides input / output operations for the system 1100. In some implementations, the input / output device 1140 may include one or more of a network interface devices, e.g., an Ethernet card, a serial communication device, e.g., an RS-232 port, and / or a wireless interface device, e.g., an 802. 11 card, a wireless modem (e.g., 3G, 4G, or 5G). In some implementations, the input / output device may include driver devices configured to receive input data and send output data to other input / output devices, e.g., keyboard, printer and display devices 1160. In some examples, mobile computing devices, mobile communication devices, and other devices may be used.

[0093] In some implementations, at least a portion of the approaches described above may be realized by instructions that upon execution cause one or more processing devices to carry out the processes and functions described above. Such instructions may include, for example, interpreted instructions such as script instructions, or executable code, or other instructions stored in a non-transitory computer readable medium. The storage device 1130 may be implemented in a distributed way over a network, for example as a server farm or a set of widely distributed servers, or may be implemented in a single computing device.

[0094] Although an example processing system has been described in FIG. 11, embodiments of the subject matter, functional operations and processes described in this specification can be implemented in other types of digital electronic circuitry, in tangibly- embodied computer software or firmware, in computer hardware, including the structures24IPTS / 200149918Docket No. KRN-001WO disclosed in this specification and their structural equivalents, or in combinations of one or more of them. Embodiments of the subject matter described in this specification can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a tangible nonvolatile program carrier for execution by, or to control the operation of, data processing apparatus. Alternatively or in addition, the program instructions can be encoded on an artificially generated propagated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal that is generated to encode information for transmission to suitable receiver apparatus for execution by a data processing apparatus. The computer storage medium can be a machine-readable storage device, a machine -readable storage substrate, a random or serial access memory device, or a combination of one or more of them.

[0095] The term “system” may encompass all kinds of apparatus, devices, and machines for processing data, including by way of example a programmable processor, a computer, or multiple processors or computers. A processing system may include special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit). A processing system may include, in addition to hardware, code that creates an execution environment for the computer program in question, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them.

[0096] A computer program (which may also be referred to or described as a program, software, a software application, an engine, a pipeline, a module, a software module, a script, or code) can be written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages, and it can be deployed in any form, including as a standalone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program may, but need not, correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub programs, or portions of code). A computer program can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a communication network.

[0097] The processes and logic flows described in this specification can be performed by one or more programmable computers executing one or more computer programs to perform25IPTS / 200149918Docket No. KRN-001WO functions by operating on input data and generating output. The processes and logic flows can also be performed by, and apparatus can also be implemented as, special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit).

[0098] Computers suitable for the execution of a computer program can include, by way of example, general or special purpose microprocessors or both, or any other kind of central processing unit. Generally, a central processing unit will receive instructions and data from a read-only memory or a random access memory or both. A computer generally includes a central processing unit for performing or executing instructions and one or more memory devices for storing instructions and data. Generally, a computer will also include, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto optical disks, or optical disks. However, a computer need not have such devices. Moreover, a computer can be embedded in another device, e.g., a mobile telephone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a Global Positioning System (GPS) receiver, or a portable storage device (e.g., a universal serial bus (USB) flash drive), to name just a few.

[0099] Computer readable media suitable for storing computer program instructions and data include all forms of nonvolatile memory, media and memory devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto optical disks; and CD-ROM and DVD-ROM disks. The processor and the memory can be supplemented by, or incorporated in, special purpose logic circuitry.

[0100] To provide for interaction with a user, embodiments of the subject matter described in this specification can be implemented on a computer having a display device, e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor, for displaying information to the user and a keyboard and a pointing device, e.g., a mouse or a trackball, by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback, e.g., visual feedback, auditory feedback, or tactile feedback; and input from the user can be received in any form, including acoustic, speech, or tactile input. In addition, a computer can interact with a user by sending documents to and receiving documents from a device that is used by the user; for example, by sending web pages to a web browser on a user’s user device in response to requests received from the web browser.26IPTS / 200149918Docket No. KRN-001WO

[0101] Embodiments of the subject matter described in this specification can be implemented in a computing system that includes a back end component, e.g., as a data server, or that includes a middleware component, e.g., an application server, or that includes a front end component, e.g., a client computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the subject matter described in this specification, or any combination of one or more such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication, e.g., a communication network. Examples of communication networks include a local area network (“LAN”) and a wide area network (“WAN”), e.g., the Internet.

[0102] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

[0103] While this specification contains many specific implementation details, these should not be construed as limitations on the scope of what may be claimed, but rather as descriptions of features that may be specific to particular embodiments. Certain features that are described in this specification in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Moreover, although features may be described above as acting in certain combinations and even initially claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination may be directed to a subcombination or variation of a subcombination.

[0104] Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Moreover, the separation of various system components in the embodiments described above should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can27IPTS / 200149918Docket No. KRN-001WO generally be integrated together in a single software product or packaged into multiple software products.

[0105] Particular embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims. For example, the actions recited in the claims can be performed in a different order and still achieve desirable results. As one example, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In certain implementations, multitasking and parallel processing may be advantageous. Other steps or stages may be provided, or steps or stages may be eliminated, from the described processes. Accordingly, other implementations are within the scope of the following claims.Terminology

[0106] The phraseology and terminology used herein is for the purpose of description and should not be regarded as limiting.

[0107] The term “approximately”, the phrase “approximately equal to”, and other similar phrases, as used in the specification and the claims (e.g., “X has a value of approximately Y” or “X is approximately equal to Y”), should be understood to mean that one value (X) is within a predetermined range of another value (Y). The predetermined range may be plus or minus 20%, 10%, 5%, 3%, 1%, 0.1%, or less than 0.1%, unless otherwise indicated.

[0108] Measurements, sizes, amounts, etc. may be presented herein in a range format. The description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the invention. Accordingly, the description of a range should be considered to have specifically disclosed all the possible subranges as well as individual numerical values within that range. For example, description of a range such as 10-20 inches should be considered to have specifically disclosed subranges such as 10-11 inches, 10-12 inches, 10-13 inches, 10-14 inches, 11-12 inches, 11-13 inches, etc.

[0109] The indefinite articles “a” and “an,” as used in the specification and in the claims, unless clearly indicated to the contrary, should be understood to mean “at least one.” The phrase “and / or,” as used in the specification and in the claims, should be understood to mean “either or both” of the elements so conjoined, i.e., elements that are conjunctively present in some cases and disjunctively present in other cases. Multiple elements listed with “and / or” should be construed in the same fashion, i.e., “one or more” of the elements so conjoined. Other elements may optionally be present other than the elements specifically identified by28IPTS / 200149918Docket No. KRN-OOIWO the “and / or” clause, whether related or unrelated to those elements specifically identified. Thus, as a non-limiting example, a reference to “A and / or B”, when used in conjunction with open-ended language such as “comprising” can refer, in one embodiment, to A only (optionally including elements other than B); in another embodiment, to B only (optionally including elements other than A); in yet another embodiment, to both A and B (optionally including other elements); etc.

[0110] As used in the specification and in the claims, “or” should be understood to have the same meaning as “and / or” as defined above. For example, when separating items in a list, “or” or “and / or” shall be interpreted as being inclusive, i.e., the inclusion of at least one, but also including more than one, of a number or list of elements, and, optionally, additional unlisted items. Only terms clearly indicated to the contrary, such as “only one of or “exactly one of,” or, when used in the claims, “consisting of,” will refer to the inclusion of exactly one element of a number or list of elements. In general, the term “or” as used shall only be interpreted as indicating exclusive alternatives (i.e. “one or the other but not both”) when preceded by terms of exclusivity, such as “either,” “one of,” “only one of,” or “exactly one of.” “Consisting essentially of,” when used in the claims, shall have its ordinary meaning as used in the field of patent law.[oni] As used in the specification and in the claims, the phrase “at least one,” in reference to a list of one or more elements, should be understood to mean at least one element selected from any one or more of the elements in the list of elements, but not necessarily including at least one of each and every element specifically listed within the list of elements and not excluding any combinations of elements in the list of elements. This definition also allows that elements may optionally be present other than the elements specifically identified within the list of elements to which the phrase “at least one” refers, whether related or unrelated to those elements specifically identified. Thus, as a non-limiting example, “at least one of A and B” (or, equivalently, “at least one of A or B,” or, equivalently “at least one of A and / or B”) can refer, in one embodiment, to at least one, optionally including more than one, A, with no B present (and optionally including elements other than B); in another embodiment, to at least one, optionally including more than one, B, with no A present (and optionally including elements other than A); in yet another embodiment, to at least one, optionally including more than one, A, and at least one, optionally including more than one, B (and optionally including other elements); etc.29IPTS / 200149918Docket No. KRN-001WO

[0112] The use of “including,” “comprising,” “having,” “containing,” “involving,” and variations thereof, is meant to encompass the items listed thereafter and additional items.

[0113] Use of ordinal terms such as “first,” “second,” “third,” etc., in the claims to modify a claim element does not by itself connote any priority, precedence, or order of one claim element over another or the temporal order in which acts of a method are performed. Ordinal terms are used merely as labels to distinguish one claim element having a certain name from another element having a same name (but for use of the ordinal term), to distinguish the claim elements.

[0114] While example embodiments have been particularly shown and described, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the embodiments encompassed by the appended claims.

[0115] What is claimed is:30IPTS / 200149918

Claims

Docket No. KRN-OOIWOCLAIMS1. A computer-implemented method comprising: providing access to a construction management platform; ingesting documents related to a construction project into the construction management platform; generating, using the construction management platform, a knowledge graph based on the ingested documents, the knowledge graph defining relationships among entities associated with the construction project; obtaining, using the construction management platform, information related to at least one external factor comprising one or more of weather, news events, or market conditions; determining, using the knowledge graph, a risk that the at least one external factor will cause a delay to a schedule for the construction project; and taking, by at least one automated agent associated with the construction management platform, a corrective action in response to the determined risk, the corrective action comprising at least one of sending a message to a stakeholder for the construction project, modifying an order for a material or service associated with the construction project, or revising the schedule for the construction project.

2. The method of claim 1, wherein the documents include at least one of a material list, a schedule, a submittal, a drawing, a pre-task plan, an order, a schematic, a request for information, a table, a bill of materials, a job hazard description, or any combination thereof.

3. The method of claim 1, wherein ingesting the documents comprises: partitioning the documents into regions; extracting metadata from the documents; performing an automated visual scan of the documents; and extracting text from the documents.

4. The method of claim 1, wherein generating the knowledge graph comprises: providing data extracted from the documents to a predictive model that generates contextualized, vector representations of the data; and constructing the knowledge graph from the contextualized, vector representations.31IPTS / 200149918Docket No. KRN-OOIWO5. The method of claim 1, wherein the knowledge graph is constructed using a material ontology that defines relationships between construction materials and at least one of a construction activity, a construction location, a materials supplier, a materials specification, or a submittal.

6. The method of claim 1, wherein the entities comprise at least one of materials, orders, schedules, activities, or prerequisites.

7. The method of claim 1, wherein taking the corrective action comprises using a voice agent to make an autonomous voice call.

8. The method of claim 1, further comprising using a graphical user interface in the construction management platform to build a workflow for the construction project, the workflow defining actions to be performed by the at least one automated agent.

9. The method of claim 8, wherein the actions include at least one of checking an inventory of a construction material, placing or modifying an order for the construction material, preparing a purchase order, making a phone call to a supplier, or sending an email or text message to a supplier.

10. A system, comprising: one or more computer processors programmed to perform operations comprising: providing access to a construction management platform; ingesting documents related to a construction project into the construction management platform; generating, using the construction management platform, a knowledge graph based on the ingested documents, the knowledge graph defining relationships among entities associated with the construction project; obtaining, using the construction management platform, information related to at least one external factor comprising one or more of weather, news events, or market conditions; determining, using the knowledge graph, a risk that the at least one external factor will cause a delay to a schedule for the construction project; and32IPTS / 200149918Docket No. KRN-001WO taking, by at least one automated agent associated with the construction management platform, a corrective action in response to the determined risk, the corrective action comprising at least one of sending a message to a stakeholder for the construction project, modifying an order for a material or service associated with the construction project, or revising the schedule for the construction project.

11. The system of claim 10, wherein the documents include at least one of a material list, a schedule, a submittal, a drawing, a pre-task plan, an order, a schematic, a request for information, a table, a bill of materials, a job hazard description, or any combination thereof.

12. The system of claim 10, wherein ingesting the documents comprises: partitioning the documents into regions; extracting metadata from the documents; performing an automated visual scan of the documents; and extracting text from the documents.

13. The system of claim 10, wherein generating the knowledge graph comprises: providing data extracted from the documents to a predictive model that generates contextualized, vector representations of the data; and constructing the knowledge graph from the contextualized, vector representations.

14. The system of claim 10, wherein the knowledge graph is constructed using a material ontology that defines relationships between construction materials and at least one of a construction activity, a construction location, a materials supplier, a materials specification, or a submittal.

15. The system of claim 10, wherein the entities comprise at least one of materials, orders, schedules, activities, or prerequisites.

16. The system of claim 10, wherein taking the corrective action comprises using a voice agent to make an autonomous voice call.33IPTS / 200149918Docket No. KRN-OOIWO17. The system of claim 10, the operations further comprising using a graphical user interface in the construction management platform to build a workflow for the construction project, the workflow defining actions to be performed by the at least one automated agent.

18. The system of claim 17, wherein the actions include at least one of checking an inventory of a construction material, placing or modifying an order for the construction material, preparing a purchase order, making a phone call to a supplier, or sending an email or text message to a supplier.

19. An article, comprising: a non-transitory computer-readable medium having instructions stored thereon that, when executed by one or more computer processors, cause the one or more computer processors to perform operations comprising: providing access to a construction management platform; ingesting documents related to a construction project into the construction management platform; generating, using the construction management platform, a knowledge graph based on the ingested documents, the knowledge graph defining relationships among entities associated with the construction project; obtaining, using the construction management platform, information related to at least one external factor comprising one or more of weather, news events, or market conditions; determining, using the knowledge graph, a risk that the at least one external factor will cause a delay to a schedule for the construction project; and taking, by at least one automated agent associated with the construction management platform, a corrective action in response to the determined risk, the corrective action comprising at least one of sending a message to a stakeholder for the construction project, modifying an order for a material or service associated with the construction project, or revising the schedule for the construction project.34IPTS / 200149918

Images