[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117900-en":3,"doc-seo-117900-105":29,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":11,"language":21,"language_code":22,"site_id":23,"html_lang":22,"table_of_contents":24,"faqs":25,"seo_title":26,"seo_description":14,"update_tm":27,"read_time":28},117900,8796095360427,"Lucas Martin","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",6,"Technology","CONSTRAINED MACHINE LEARNING MODEL DEPLOYMENTS FOR OPERATIONAL TECHNOLOGY NETWORKS","Techniques are presented for formulating level- and device-specific machine learning (ML) models tailored to operational technology (OT) networks. The approach targets constrained devices by enabling deployment closer to end devices within each Purdue model level. It addresses how different OT levels use different Ethernet protocols, which changes the data and the ML inference tasks that should run at each location. The method distributes edge ML computing while preserving the Purdue structure and avoiding violations of the security model.","Technical Disclosure Commons  \nDefensive Publications Series  \nMay 2023  \nCONSTRAINED MACHINE LEARNING MODEL DEPLOYMENTS FOR OPERATIONAL TECHNOLOGY NETWORKS  \nRobert Barton  \nIndermeet Gandhi Jerome Henry  \nFollow this and additional works at: [https://www.tdcommons.org/dpubs_series](https://www.tdcommons.org/dpubs_series)  \nRecommended Citation  \nBarton, Robert; Gandhi, Indermeet; and Henry, Jerome, \"CONSTRAINED MACHINE LEARNING MODEL DEPLOYMENTS FOR OPERATIONAL TECHNOLOGY NETWORKS\", Technical Disclosure Commons,(May 18, 2023)  \n[https://www.tdcommons.org/dpubs_series/5911](https://www.tdcommons.org/dpubs_series/5911)  \nThis work is licensed under a Creative Commons Attribution 4.0 License.  \nThis Article is brought to you for free and open access by Technical Disclosure Commons. It has been accepted for inclusion in Defensive Publications Series by an authorized administrator of Technical Disclosure Commons.  \nCONSTRAINED MACHINE LEARNING MODEL DEPLOYMENTS FOR  \nOPERATIONAL TECHNOLOGY NETWORKS  \nAUTHORS:  \nRobert Barton  \nIndermeet Gandhi  \nJerome Henry  \nABSTRACT  \nPresented herein are techniques for formulating level and device specific machine learning (ML) models for operational technology (OT) networks that can be deployed closer to an end device (in a respective level) for constrained devices.  \nDETAILED DESCRIPTION  \nOperational technology (OT) industrial networks, such as Industrial Internet of Things (IIoT) networks are organized into different levels. These levels range from Level 0 (Process Control) to Level 4 (Enterprise) and each level is responsible for a specific function in an IIoT environment. The industrial Ethernet (IE) switches deployed at each level facilitate communication between different OT devices within the same level.  \nIn Level 0, the IE switches connect to various OT devices such as sensors, actuators, and programmable logic controllers (PLCs) . These devices are responsible for controlling the physical processes in the IIoT environment. The IE switches at Level 0 typically use proprietary protocols to communicate with the OT devices.  \nAt Level 1, the IE switches connect to the controllers that manage the physical processes in the IIoT environment. These controllers include Distributed Control Systems (DCS), Supervisory Control and Data Acquisition (SCADA) systems, Programmable Automation Controllers (PAC), and the like. The IE switches at Level 1 typically use standard Ethernet protocols such as Modbus, OPC, and Ethernet/IP to communicate with the controllers.  \nIn Level 2, the IE switches connect to devices responsible for managing production processes, such as Manufacturing Execution Systems (MES) and Human-Machine Interfaces (HMI) . The IE switches at Level 2 use standard Ethernet protocols such as TCP/IP and [HTTP to communicate with these devices](HTTP to communicate with these devices).  \n1 6886  \nPublished by Technical Disclosure Commons, 2023 2  \nIn Level 3, the IE switches connect to devices responsible for managing the data collected from the OT devices in the lower levels. These devices can include Historians, Data Warehouses, and Analytics engines. The IE switches at Level 3 use standard Ethernet protocols such as Message Queues Telemetry Transport (MQTT), Advanced Message Queuing Protocol (AMQP), and Representational State Transfer (RESTful) Application Programming Interfaces (APIs) to communicate with these devices.  \nFinally, in Level 4, the IE switches connect to the devices responsible for managing the business processes in the IIoT environment. These devices include Enterprise Resource Planning (ERP) systems, Customer Relationship Management (CRM) systems, and Supply Chain Management (SCM) systems. The IE switches at Level 4 use standard Ethernet protocols such as Simple Object Access Protocol (SOAP) and Extensible Markup Language-Remote Procedure Call (XML-RPC) to communicate with these devices.  \nThe constrained IE switches deployed at each Purdue model can be used to run d","cbCaiqFEhCAiQUMg","https://ap.wps.com/l/cbCaiqFEhCAiQUMg","pdf",127476,1,"English","en",105,"# Overview\n## OT network levels and Purdue model\n## Constrained edge computing and security constraints\n# Level- and device-specific ML model deployment","[{\"question\":\"What problem does the proposal address for edge machine learning in OT networks?\",\"answer\":\"It addresses the need to distribute ML computing at the edge without violating the Purdue structure, which can be breached when edge resources are insufficient.\"},{\"question\":\"How does the Purdue model influence ML deployment in OT environments?\",\"answer\":\"Each Purdue level corresponds to specific OT functions and communication protocols, so the data and the ML tasks suitable for inference differ by level.\"},{\"question\":\"What types of constraints are considered in deploying ML closer to end devices?\",\"answer\":\"The document focuses on constrained edge resources, such as limited compute capacity (e.g., lacking GPUs), which affects where training or inference can run.\"}]","CONSTRAINED MACHINE LEARNING MODEL DEPLOYMENTS FOR OPERATIONAL TECHNOLOGY NETWORKS | PDF",1785680275,15,{"code":4,"msg":30,"data":31},"ok",{"site_id":23,"language":22,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"constrained-machine-learning-model-deployments-for-operational-technology-networks","",{"@graph":35,"@context":84},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/technology/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/constrained-machine-learning-model-deployments-for-operational-technology-networks/117900/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":22,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-08-02",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What problem does the proposal address for edge machine learning in OT networks?","Question",{"text":74,"@type":75},"It addresses the need to distribute ML computing at the edge without violating the Purdue structure, which can be breached when edge resources are insufficient.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"How does the Purdue model influence ML deployment in OT environments?",{"text":79,"@type":75},"Each Purdue level corresponds to specific OT functions and communication protocols, so the data and the ML tasks suitable for inference differ by level.",{"name":81,"@type":72,"acceptedAnswer":82},"What types of constraints are considered in deploying ML closer to end devices?",{"text":83,"@type":75},"The document focuses on constrained edge resources, such as limited compute capacity (e.g., lacking GPUs), which affects where training or inference can run.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":23},{"code":4,"msg":5,"data":91},[92,96,100,104,109,112,117,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":110,"slug":111},50,"technology",{"id":113,"doc_module":4,"doc_module_name":45,"category_name":114,"show_sort_weight":115,"slug":116},7,"Healthcare",40,"healthcare",{"id":118,"doc_module":4,"doc_module_name":45,"category_name":119,"show_sort_weight":120,"slug":121},8,"Research & Report",30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":105,"slug":137},19,"General","general"]