LogoLogo
IntegrationsInstallationAdministrationContact Support
XMPro Platform
XMPro Platform
  • What is XMPro?
  • Getting Started
    • Browser Requirements
    • Free Trial
    • End-To-End Use Case
  • Resources
    • What's New in 4.4
      • What's New in 4.3
      • What's New in 4.2
      • What's New in 4.1.13
      • What's New in 4.1
      • What's New in 4.0
    • Blueprints, Accelerators & Patterns
    • Integrations
    • Sizing Guideline
    • Platform Security
    • Icon Library
    • FAQs
      • Implementation FAQs
      • Configuration FAQs
      • Agent FAQs
      • General FAQs
      • External Content
        • Blogs
          • 2024
            • How to Build Multi-Agent Systems for Industry
            • Why Solving the Problem Doesn’t Solve the Problem: The Importance of Scalable Intelligent Operations
            • Content, Decision, and Hybrid: The Three Pillars of Multi-Agent Systems in Industry
            • Revolutionizing Manufacturing with AI and Generative AI: XMPro’s Intelligent Business Operations Sui
            • The Evolution of Skills: Lessons from Agriculture in the GenAI and MAGS Era
            • Part 1: From Railroads to AI: The Evolution of Game-Changing Utilities
            • Part2: The Future of Work: Harnessing Generative Agents in Manufacturing
            • Bridging Automation and Intelligence: XMPro’s Approach to Industrial Agent Management
            • XMPro APEX: Pioneering AgentOps for Industrial Multi Agent Generative Systems
            • Part 5 – Rules of Engagement: Establishing Governance for Multi-Agent Generative Systems
            • How to Achieve Scalable Predictive Maintenance for Industrial Operations
            • Understanding the Difference Between XMPro AI Assistant and AI Advisor
            • Part 3 – AI at the Core: LLMs and Data Pipelines for Industrial Multi-Agent Generative Systems
            • MAGS: The Killer App for Generative AI in Industrial Applications
            • The Importance of Pump Predictive Maintenance for Operational Efficiency
            • Progressing Through The Decision Intelligence Continuum With XMPro
            • The Value-First Approach to Industrial AI: Why MAGS Implementation Must Start with Business Outcomes
            • New Guide – The Ultimate Guide to Multi-Agent Generative Systems
            • The Ultimate Guide To Predictive Analytics
            • Part 4 – Pioneering Progress | Real-World Applications of Multi-Agent Generative Systems
            • Scaling Multi-Agent Systems with Data Pipelines: Solving Real-World Industrial Challenges
          • 2023
            • How to master Predictive Analytics using Composable Digital Twins
            • Accelerate Your AI Workflow: The 3 Key Business Advantages of XMPro Notebook
            • The Roadmap to Intelligent Digital Twins
            • What is edge computing, and how can digital twins utilize this technology?
            • THE TOP 5 USE CASES FOR COMPOSABLE DIGITAL TWINS IN RENEWABLES + HOW TO SUPERCHARGE RESULTS WITH AI
            • The Technology Behind Predictive Maintenance (PdM) : Hardware & Software
            • The Benefits of Using Digital Twins in Smart Manufacturing
            • XMPro I3C Intelligent Digital Twins Strategy Framework
            • The TOP 5 use cases for composable digital twins in mining – and how to use AI to supercharge result
            • The TOP 5 use cases for Composable Digital Twins in the Oil & Gas industry
            • Why Decision Intelligence with Digital Twins is “kinda like” DCS for Automation and Control
            • XMPro becomes an NVIDIA Cloud-Validated partner
            • From Reactive to Predictive : Introduction to Predictive Maintenance
            • Microsoft Azure Digital Twins : Everything You Need To Know
            • Unlocking Efficiency: The Right Time & Strategy to Launch Your Digital Twin for Enhanced Asset Manag
            • Revolutionize Your Supply Chain: How Digital Twins Can Boost Efficiency and Cut Costs
          • 2022
            • Create a Common Operating Picture of Your Operations with XMPro
            • 7 Trends for Industrial Digital Twins in 2022
            • How to Build a Digital Twin + 60 Use Cases By Industry
            • What are composable digital twins in the metaverse?
          • 2021
            • The Value of a Composable Digital Twin
          • 2020
            • Lean Digital Twin: Part 2
            • Digital Twin: Your Most Productive Remote Worker
            • From the Control Room to the Bedroom
            • Lean Digital Twin: Part 3
          • 2019
            • My Digital Twin: Digital Twin Applications for Real-time Operations (Like Me)
          • 2018
            • XMPro IoT Operational Capability Survey Results 2018
            • What is a Digital Business Platform and Why Should I Care?
            • [Robotic] Process Automation for IoT
            • 3 Patterns of Industrial IoT Use Cases
            • The CXO’s Guide to Digital Transformation – May The Five Forces Be With You
            • Is Security More Important Than Trustworthiness for Industrial IoT?
            • XMPro at bpmNEXT 2018: Watch The Presentation
          • 2017
            • The Top 5 Reasons to Invest in an IIoT Development Platform
            • IoT Business Solutions Start with Big Data & Create Business Outcomes
            • How AI Bots Bring Digital Twins to Life
          • 2016
            • How To Get Started With Industrial IoT
            • How To Overcome The Top 5 Challenges To Industrial IoT Adoption
            • What is an IoT Platform vs. an IoT Business Application Suite?
            • Industrial IoT: How To Get Started with Predictive Maintenance
            • 3 Ways The Internet of Things is Transforming Field Service
            • 7 Types of Industrial IoT Data Sources (And How To Use Them)
          • 2015
            • How Important Are Processes To The Internet Of Things?
            • Understanding the Value of Real Time KPI Management as Your Next Strategic Project
            • 6 Myths About Machine Learning
            • 10 Predictive Analytics Use Cases By Industry
            • What is a “Business Moment” in your business?
            • Does Operational Intelligence Make Business Intelligence Obsolete?
            • How To Reduce Operational Costs by 36% with Predictive Analytics
            • From Many, One – The Nature of Complex Event Processing
            • Herding Cats: What Enterprise Architects need to know about Business Process Management
          • 2014
            • Making Business Operations More Intelligent
          • 2013
            • Best Next Action Is The Next Big Thing For Intelligent Operations
            • The learns from two ‘Best in class’ organisations acquiring BPM technology
          • 2012
            • Why Intelligent Business Operations is Mobile, Social and Smart
            • Why Do You Want Intelligent Business Operations?
            • How big of a problem are ‘dark processes’?
            • Operational Risk: When You Stick Your Head In The Sand
            • The Difference Between Event-based And Workflow-based Processes
          • 2011
            • Is mobile BPM now essential to the business?
            • Stretch Socially Dynamic Processes To Fit Your Business
            • Social Listening – Get Control Of The Conversation
            • Operations Management – The Keys To KPIs
            • Benefits of BPM v 1.0
            • How to Prioritise Processes
          • 2010
            • The Business Drivers
            • Preserving Capability and Agility
            • Mobile BPM
        • Use Cases
          • Aging Pipe Predictive Maintenance in Water Utilities
          • Air Quality Monitoring For Agriculture
          • Alarm Management and Triage
          • Asset Condition Monitoring for Surface Processing Plants in the Mining Industry
          • Bogie Health Monitoring in the Rail Industry
          • Boiler Feed Water Pumps
          • CHPP Throughput Loss Monitoring
          • Casting Guidance
          • Conveyor Belt System Monitoring and Optimization in Automotive Manufacturing
          • Cooling Tower Fin Fan Monitoring
          • Cyclone/Slurry Pump Monitoring
          • Demand Planning to Reduce Stockholding in Stores
          • Demin Water Monitoring for Boiler Tube Corrosion
          • EV Battery Assembly Process Optimization for the Car Manufacturing Industry
          • Flood Prediction & Response in Water Utilities
          • Golden Batch For Culture Addition In The Dairy processing Industry.
          • Golden Batch Monitoring
          • Improve First Pass Yield (FPY)
          • Induced Draft (ID) Fan Monitoring
          • Long Conveyor Monitoring
          • Monitor Process Health to Reduce Cash-to-Cash Cycle
          • Monitor Storm Water Reservoirs For Flood Prevention
          • Monitor and Reduce Energy Consumption
          • Oil Well Maintenance Planning
          • Oil Well RTP Monitoring
          • Pipe Scaling Prediction for Roller Cooling
          • Precision Irrigation in Agriculture
          • Predict Heat Exchanger Fouling
          • Predictive Maintenance & Asset Health Monitoring For Haul Trucks In The Mining Industry
          • Predictive Maintenance For Mobile Assets Within The Mining Industry
          • Predictive Maintenance for Robotic Arms in the Automotive Industry
          • Predictive Maintenance for Wind Turbines
          • Pump Health Monitoring in Water Utilities
          • Pumping Station OEE
          • Real-time Balanced Business Scorecard (BBS)
          • Real-time Safety Monitoring
          • Short Term Inventory Planning
          • Strategic Performance & Safety Oversight for Global Mining Operations
          • Wheel and Track Wear Monitoring In The Rail Industry
          • Wind Turbine Performance Optimization
        • Youtube
          • 2024
            • Discover Gen AI Powered Operations With XMPro iBOS
            • Generative AI and Digital Twins in 2024 - XMPro Webinar
            • Go From Reactive To Predictive Operations In Water Utilities With XMPro iDTS
            • How to add Timestamps to Elements in XMPro App Designer
            • How to Build an AI Advisor for Industrial Operations Using XMPro
            • How XMPro Stream Hosts and Collections Enable Scalable, Real-Time Data Processing
            • Mind Blowing AI Agentic Operations For Industry With XMPro MAGS
            • The Ultimate Beginner's Guide To Predictive Analytics Podcast
            • XMPro's Flexible Deployment Options: Flexible Cloud & On-Premise Solutions For Industry
            • XMPro iBOS: The Only AI-Powered Suite for Scalable Intelligent Operations
          • 2023
            • 2023 XMPro Product Roadmap - Webinar
            • An Introduction To Intelligent Digital Twins - Webinar
            • Energy and Utilities Asset Optimisation through Digital Twin technology
            • Explore Model Governance using our MLflow Agent
            • Exploring XMPro Notebook and MLflow for Data Science and Model Governance
            • How Changing Properties For One Block Can Be Applied To All Blocks Within Same Style Group
            • How do I Use A Button To Update a Data Source In XMPro App Designer
            • How Does XMPro Compare To ESBs (Enterprise Service Buses)-
            • How to Configure and Integrity Check in Data Streams
            • How To Create A Widget Within XMPro App Designer
            • How to Create Intelligent Digital Twins Using XMPro AI
            • How to export grid data to Excel In XMPro App Designer
            • How to Revolutionize Your Supply Chain with Digital Twins
            • How To Rotate Text In App Designer
            • How To Update a Data Source Using A Button
            • How To Use & Clone XMPro Demos For Your Own Use
            • How To Use And Build 3rd Party Apps To Extend The Capabilities Of The XMPro App Designer.
            • How to use Avatars and why they are important
            • How to view stream host logs In XMPro Data Stream Designer
            • Logging Provider Support With XMPro
            • Mastering Health Check Endpoints: A Guide to Ensuring Service Uptime and Performance with XMPro
            • Mastering Root Cause Analysis with XMPro: Capture, Value, Impact
            • Microsoft Azure Digital Twins Everything You Need To Know
            • Model Based Predictive Maintenance (PdM) With XMPro
            • Monthly Webinar - Accelerate your digital twin use cases - XMPro Blueprints, Accelerators & Patterns
            • Optimizing Time Series Chart (TSC) Performance
            • Predictive Maintenance & Condition Monitoring - A Hot Seat Q&A Session
            • Predictive Maintenance with XMPro iDTS
            • Smart Facilities Management with Intelligent Digital Twins
            • The Benefits of using Digital Twins in Smart Manufacturing
            • The Four Industrial Revolutions Explained In Under 4 Minutes! #industry4 #smartmanufacturing
            • The Roadmap To Intelligent Digital Twins
            • The Technology Behind Predictive Maintenance (PdM) - The Hardware & Software that makes PdM Tick...
            • THE TOP USE CASES FOR COMPOSABLE DIGITAL TWINS IN RENEWABLES
            • Tips on how to use cache in agent configuration and get live updates
            • Webinar - XMPro 4.3 Release Showcase
            • What is a Digital Twin- Why Composable Digital Twins is the Future.
            • What Is Predictive Maintenance- (PdM)
            • What To Do When a Data Source Is Not Showing in Pass Page Parameter
            • XMPro - The World's Only AI - Powered Intelligent Digital Twin Suite
            • XMPro - The World's Only No Code Digital Twin Composition Platform
            • XMPro AI : How It Works
            • XMPro AI End To End Use Case
            • XMPro Auto Scale - Understanding Distributed Caching for Cloud-Native Applications
            • XMPro Promo Video - Dell Validated Design For Manufacturing Edge
          • 2022
            • Aggregate Transformation Agent Example - XMPRO Data Stream Designer
            • App Layout Best Practices for Desktop & Mobile - XMPro Lunch & Learn
            • Broadcast Transformation Agent Example - XMPRO Data Stream Designer
            • Calculated Field Transformation Agent Example - XMPRO Data Stream Designer
            • CRC16 Function Agent Example - XMPRO Data Stream Designer
            • Create a Common Operating Picture of Your Operations with XMPro
            • CSV Context Provider Agent Example - XMPro Data Stream Designer
            • CSV Simulator Agent Example - XMPRO Data Stream Designer
            • CSV Writer Agent Example - XMPRO Data Stream Designer
            • Data Conversion Transformation Agent Example - XMPro Data Stream Designer
            • Digital Twin Strategy To Execution Pyramid - XMPro Webinar
            • Event Printer Action Agent Example - XMPRO Data Stream Designer
            • File Listener Agent Example - XMPRO Data Stream Designer
            • Filter Transformation Agent Example - XMPRO Data Stream Designer
            • Group & Merge Transformation Agent Example - XMPRO Data Stream Designer
            • How To Bind Data To A Chart and Get It Working As Expected - XMPro Lunch & Learn
            • How To Send Data To My App (Including Caching Introduction) - XMPro Lunch & Learn
            • Join Transformation Agent Example - XMPRO Data Stream Designer
            • Min/Max Function Agent Example - XMPRO Data Stream Designer
            • PART 1- How To Manage Complex Operations in Real-time Using Composable Digital Twins
            • PART 3 - How To Manage Complex Operations in Real-time Using Composable Digital Twins
            • PART2 - How To Manage Complex Operations in Real-time Using Composable Digital Twins
            • Pass Through Agent Example - XMPRO Data Stream Designer
            • Pivot Table Transformation Agent Example - Count - XMPRO Data Stream Designer
            • Pivot Table Transformation Agent Example - Sum - XMPRO Data Stream Designer
            • Real-Time Is Real - How To Use Event Intelligence Tools to Manage Complex Operations in Real-time.
            • Row Count Agent Example - XMPRO Data Stream Designer
            • Sort Transformation Agent Example - XMPRO Data Stream Designer
            • Transpose Transformation Agent Example - Columns - XMPRO Data Stream Designer
            • Transpose Transformation Agent Example - Rows - XMPRO Data Stream Designer
            • Trim Name Transformation Agent Example - XMPRO Data Stream Designer
            • Twilio Action Agent Example - XMPRO Data Stream Designer
            • Union Transformation Agent Example - XMPRO Data Stream Designer
            • Variables & Expressions in App Designer - XMPro Lunch & Learn
            • Window Transformation Agent Example - XMPRO Data Stream Designer
            • XML File Reader Action Agent Example - XMPRO Data Stream Designer
          • 2021
            • The Value of a Composable Digital Twin - XMPro Webinar
          • 2020
            • 1. Understanding The Problem - UX Design - XMPRO
            • 1.1 Welcome - XMPRO UI Design Basics
            • 1.2 Introduction To UI Design - XMPRO UI Design Basics
            • 2. Creating User Stories - UX Design - XMPRO
            • 2.1 Responsive Design - XMPRO UI Design Basics
            • 2.2 Grids - XMPRO UI Design Basics
            • 2.3 Visual Hierarchy - XMPRO UI Design Basics
            • 2.4 Wireframes - XMPRO UI Design Basics
            • 3. Creating User Flow Diagrams - UX Design - XMPRO
            • 3.1 Color Palette - XMPRO UI Design Basics
            • 3.2 Typography - XMPRO UI Design Basics
            • 3.3 White Space - XMPRO UI Design Basics
            • 3.4 UI Elements - XMPRO UI Design Basics
            • 4. Plan Your App with Wireframes - UX Design - XMPRO
            • 4.1 Chart Types - XMPRO UI Design Basics
            • 4.2 Chart Styling - XMPRO UI Design Basics
            • 5. Designing for Dynamic Data - UX Design - XMPRO
            • Agents and Their Types - XMPRO Data Stream Designer
            • Data Wrangling: Row Transpose - XMPRO Data Stream Designer
            • Digital Twin: Your Most Productive Remote Worker - XMPRO Webinar
            • End-To-End Real-Time Condition Monitoring Demo - XMPRO Application Development Platform
            • Error Endpoints - XMPRO Data Stream Designer
            • Export and Import Recommendations - XMPRO App Designer
            • How To Add Buttons To Agents - XMPRO Data Stream Designer
            • How To Add EditLists to Agents - XMPRO Data Stream Designer
            • How To Change UI Language - XMPRO Subscription Manager
            • How To Configure a Stream Object - XMPRO Data Stream Designer
            • How To Configure The Aggregate Transformation - XMPRO Data Stream Designer
            • How To Configure The Anomaly Detection Agent - XMPRO Data Stream Designer
            • How To Configure The Azure SQL Action Agent - XMPRO Data Stream Designer
            • How To Configure The Azure SQL Context Provider - XMPRO Data Stream Designer
            • How To Configure The Azure SQL Listener - XMPRO Data Stream Designer
            • How To Configure The Calculated Field Transformation - XMPRO Data Stream Designer
            • How To Configure The CSV Context Provider - XMPRO Data Stream Designer
            • How To Configure The CSV Listener - XMPRO Data Stream Designer
            • How To Configure The Data Conversion Transformation - XMPRO Data Stream Designer
            • How To Configure The Edge Analysis Transformation - XMPRO Data Stream Designer
            • How To Configure The Email Action Agent - XMPRO Data Stream Designer
            • How To Configure The Email Listener - XMPRO Data Stream Designer
            • How To Configure The Event Printer Action Agent - XMPRO Data Stream Designer
            • How To Configure The Event Simulator Listener - XMPRO Data Stream Designer
            • How To Configure The FFT Function - XMPRO Data Stream Designer
            • How To Configure The File Listener - XMPRO Data Stream Designer
            • How To Configure The Filter Transformation - XMPRO Data Stream Designer
            • How To Configure The IBM Maximo Action Agent - XMPRO Data Stream Designer
            • How To Configure The IBM Maximo Context Provider - XMPRO Data Stream Designer
            • How To Configure The IBM Maximo Listener - XMPRO Data Stream Designer
            • How To Configure The Join Transformation - XMPRO Data Stream Designer
            • How To Configure The JSON File Reader Context Provider - XMPRO Data Stream Designer
            • How To Configure The MQTT Action Agent - XMPRO Data Stream Designer
            • How To Configure The MQTT Advanced Action Agent - XMPRO Data Stream Designer
            • How To Configure The MQTT Advanced Listener - XMPRO Data Stream Designer
            • How To Configure The MQTT Listener - XMPRO Data Stream Designer
            • How To Configure The Normalize Fields Function - XMPRO Data Stream Designer
            • How To Configure The OSIsoft PI Context Provider - XMPRO Data Stream Designer
            • How To Configure The OSIsoft PI Listener - XMPRO Data Stream Designer
            • How To Configure The Pass Through Transformation - XMPRO Data Stream Designer
            • How To Configure The PMML Agent - XMPRO Data Stream Designer
            • How To Configure The REST API Context Provider - XMPRO Data Stream Designer
            • How To Configure The RScript Agent - XMPRO Data Stream Designer
            • How To Configure The Run Recommendation Agent - XMPRO Data Stream Designer
            • How To Configure The Signal Filter - XMPRO Data Stream Designer
            • How To Configure The SQL Server Action Agent - XMPRO Data Stream Designer
            • How To Configure The SQL Server Context Provider - XMPRO Data Stream Designer
            • How To Configure The SQL Server Listener - XMPRO Data Stream Designer
            • How To Configure The SQL Server Writer Action Agent - XMPRO Data Stream Designer
            • How To Configure The Twilio Action Agent - XMPRO Data Stream Designer
            • How To Configure The Union Transformation - XMPRO Data Stream Designer
            • How To Configure The Unzip Function - XMPRO Data Stream Designer
            • How To Configure The Window Transformation - XMPRO Data Stream Designer
            • How To Create an App - XMPRO App Designer
            • How To Create and Manage Templates - XMPRO App Designer
            • How To Create and Publish a Use Case - XMPRO Data Stream Designer
            • How To Create and Use a Widget - XMPRO App Designer
            • How To Create App Data Connections - XMPRO App Designer
            • How To Create App Pages and Navigation - XMPRO App Designer
            • How To Create Recommendation Rules - XMPRO App Designer
            • How To Create Recurrent Data Streams - XMPRO Data Stream Designer
            • How To Do Integrity Checks - XMPRO Data Stream Designer
            • How To Edit Page Properties - XMPRO App Designer
            • How To Enable Audit Trails - XMPRO App Designer
            • How to Export, Import, and Clone a Data Stream - XMPRO Data Stream Designer
            • How To Export, Import and Clone an App - XMPRO App Designer
            • How to Export and Import an App - XMPRO App Designer
            • How To Find Help for an Agent - XMPRO Data Stream Designer
            • How To Install The XMPRO App Designer
            • How To Maintain and Capture Notes - XMPRO App Designer
            • How To Manage Agents - XMPRO Data Stream Designer
            • How To Manage and Use Server Variables - XMPRO Data Stream Designer
            • How To Manage Buffer Size - XMPRO Data Stream Designer
            • How to Manage Categories - XMPRO App Designer
            • How To Manage Categories - XMPRO Data Stream Designer
            • How To Pass Parameters Between Pages - XMPRO App Designer
            • How To Publish and Share an Application - XMPRO App Designer
            • How To Set Up and Use Charts in Live View - XMPRO Data Stream Designer
            • How To Set Up and Use Gauges in Live View - XMPRO Data Stream Designer
            • How To Share a Data Stream - XMPRO Data Stream Designer
            • How To Share a Use Case - XMPRO Data Stream Designer
            • How To Share an App For Design Collaboration - XMPRO App Designer
            • How To Troubleshoot a Use Case - XMPRO Data Stream Designer
            • How To Upgrade a Stream Object Version - XMPRO Data Stream Designer
            • How To Use App Files - XMPRO App Designer
            • How To Use Application Versions - XMPRO App Designer
            • How To Use Bar Gauge - XMPRO App Designer
            • How To Use Calendar - XMPRO App Designer
            • How To Use Chart Pan, Zoom and Aggregation - XMPRO App Designer
            • How To Use Chart Panes and Axes - XMPRO App Designer
            • How To Use Chart Print and Export- XMPRO App Designer
            • How To Use Charts - XMPRO App Designer Toolbox
            • How To Use Charts: Series - XMPRO App Designer
            • How To Use Collections - XMPRO Data Stream Designer
            • How To Use Content Card - XMPRO App Designer
            • How To Use D3 - XMPRO App Designer
            • How To Use Data Sources - XMPRO App Designer
            • How To Use Embedded Pages - XMPRO App Designer Toolbox
            • How To Use Fieldset and Field - XMPRO App Designer Toolbox
            • How To Use Flex Layout
            • How To Use Form Validation - XMPRO App Designer Toolbox
            • How To Use Input Mappings - XMPRO Data Stream Designer
            • How To Use Linear Gauges - XMPRO App Designer
            • How To Use Live View - XMPRO Data Stream Designer
            • How To Use Lookup - XMPRO App Designer
            • How To Use Maps - XMPRO App Designer
            • How To Use Page Layers - XMPRO App Designer
            • How To Use Pivot Grid - XMPRO App Designer
            • How To Use Polar Charts - XMPRO App Designer
            • How To Use Power BI - XMPRO App Designer
            • How To Use Radio Buttons - XMPRO App Designer Toolbox
            • How To Use Recommendations - XMPRO App Designer Toolbox
            • How To Use Select Box - XMPRO App Designer
            • How To Use Stacked Layouts - XMPRO App Designer Toolbox
            • How To Use Stream Host Local Variables - XMPRO Data Stream Designer
            • How To Use Tabs - XMPRO App Designer Toolbox
            • How To Use Tags - XMPRO App Designer Toolbox
            • How To Use Templated List - XMPRO App Designer
            • How To Use Templates - XMPRO App Designer
            • How To Use Text - XMPRO App Designer Toolbox
            • How To Use Text Area - XMPRO App Designer Toolbox
            • How To Use The Accordion - XMPRO App Designer Toolbox
            • How To Use The Block Styling Manager - XMPRO App Designer
            • How To Use The Box and Data Repeater Box - XMPRO App Designer Toolbox
            • How To Use The Button - XMPRO App Designer Toolbox
            • How To Use The Circular Gauge - XMPRO App Designer Toolbox
            • How To Use The Data Grid - XMPRO App Designer Toolbox
            • How To Use The HTML Editor - XMPRO App Designer Toolbox
            • How To Use The Hyperlink and Box Hyperlink - XMPro App Designer Toolbox
            • How To Use The Image - XMPRO App Designer Toolbox
            • How To Use The Indicator - XMPRO App Designer Toolbox
            • How To Use The Layout Grid - XMPRO App Designer Toolbox
            • How To Use The Number Selector - XMPRO App Designer Toolbox
            • How To Use The Pie Chart - XMPRO App Designer Toolbox
            • How To Use The Range Slider - XMPRO App Designer Toolbox
            • How To Use The Recommendation Chart - XMPRO App Designer Toolbox
            • How To Use The Scroll Box - XMPRO App Designer Toolbox
            • How To Use The Select Box - XMPRO App Designer Toolbox
            • How To Use The Sparkline - XMPRO App Designer Toolbox
            • How To Use The Textbox - XMPRO App Designer Toolbox
            • How To Use Tree Grid - XMPRO App Designer
            • How To Use Tree List - XMPRO App Designer
            • How To Use Unity - XMPRO App Designer Toolbox
            • How To Use Variables - XMPRO App Designer
            • How To Write and Maintain Notes and Business Case - XMPRO Data Stream Designer
            • Interactive 3D Models For Digital Twins - XMPRO Event Intelligence Platform
            • Manage Input Arrow Highlights - XMPRO Data Stream Designer
            • Manage Recommendation Access - XMPRO App Designer
            • Realize Value from End-To-End Condition Monitoring in 6 - 8 Weeks - XMPRO
            • Recommendation Versions - XMPRO App Designer
            • Solution Development Process For Event Intelligence Apps - XMPRO
            • Stream Hosts and How To Install Them - XMPRO Data Stream Designer
            • Use Case Versioning - XMPRO Data Stream Designer
            • XMPRO App Designer Overview - Event Intelligence Applications
            • XMPRO Data Stream Designer - Event Intelligence Applications
            • XMPRO Real-Time Event Intelligence Demo
            • XMPRO Recommendations - Event Intelligence Applications
          • 2019
            • Data Distribution Service: Using DDS in Your IoT Applications
            • My Digital Twin: Digital Twin Applications For Real-Time Operations (Like Me)
            • Setting up a Typical Industrial IoT Use Case with XMPro
            • XMPro Overview & Fin Fan Failure Demo
          • 2016
            • XMPro iBPMS Overview
          • 2013
            • XMPro Best Next Action - 3 Examples for XMPro blog
            • XMPro Case Management Example
            • XMPro Internet of Things Demo
          • 2012
            • Is Agile Business the New Normal
            • The Future of BPM Moving Towards Intelligent Business Operations
            • What industries does XMPro serve-
            • Who is XMPro for-
            • XMPro - The Social Listener - Why You Should Be Listening.wmv
            • XMPro Cool Vendor 2012
            • XMPro iBPMS For SharePoint
            • XMPro iBPMS v6 XMWeb for Intelligent Business Operations
            • XMPro News and Gartner BPM Sydney Summit Discount Offer.mp4
            • XMPro Version 6 - Introducing the Next Generation BPM for Intelligent Business Operations
    • Practice Notes
      • Unified Recommendation Alert Management
      • Performant Landing Pages in Real-Time Monitoring
  • Concepts
    • XMPro AI
      • XMPro Notebook
    • Data Stream
      • Stream Object Configuration
      • Verifying Stream Integrity
      • Running Data Streams
      • Timeline
    • Collection and Stream Host
    • Agent
      • Virtual vs Non-Virtual Agents
    • Application
      • Template
      • Page
      • Block
      • Canvas
      • Page Layers
      • Block Styling
      • Devices
      • Flex
      • Block Properties
      • Data Integration
      • Navigation and Parameters
      • Variables and Expressions
      • App Files
      • Metablocks
    • Recommendation
      • Rule
      • Execution Order
      • Auto Escalate
      • Form
      • Action Requests
      • Notification
      • Recommendation Alert
      • Deleted Items
      • Scoring
    • Connector
    • Landing Pages & Favorites
    • Version
    • Manage Access
    • Category
    • Variable
    • Insights
      • Data Delivery Insights
  • How-To Guides
    • Data Streams
      • Manage Data Streams
      • Manage Collections
      • Use Remote Receivers and Publishers
      • Manage Recurrent Data Streams
      • Use Business Case and Notes
      • Run an Integrity Check
      • Check Data Stream Logs
      • Use Live View
      • Use Stream Metrics
      • Troubleshoot a Data Stream
      • Upgrade a Stream Object Version
      • Setup Input Mappings
      • Use Error Endpoints
      • Use the Timeline
      • Context Menu
    • Application
      • Manage Apps
      • Manage Templates
      • Manage Pages
      • Import an App Page
      • Design Pages for Mobile
      • Navigate Between Pages
      • Pass Parameters Between Pages
      • Page Data
      • Manage Connections
      • Check Connector Logs
      • Manage Data Sources
      • Use Data Sources in the Page
      • Use Dynamic Properties
      • Use Expression Properties
      • Use Page Layers
      • Use Block Styling and Devices
      • Use Flex
      • Use Validation
      • Use Variables & Expressions
      • Create and Maintain Notes
      • Manage Widgets
      • Manage App Files
      • Manage Themes
    • Recommendations
      • Manage Categories
      • Manage Recommendations
      • Manage Rules
      • Manage Notifications
      • Manage Notification Templates
      • Subscribe to Notifications
      • Manage Forms
      • Manage Variables
      • Manage Alerts
      • Manage Alerts on Mobile
      • Manage Deleted Recommendation Items
    • Connectors
      • Manage Connectors
      • Building Connectors
      • Packaging Connectors
    • Stream Host
    • Agents
      • Manage Agents
      • Building Agents
      • Packaging Agents
      • Debugging an Agent
    • Manage Versions
    • Manage Access
    • Manage Categories
    • Manage Variables
    • Import, Export, and Clone
    • Publish
      • Admin Unpublish Override
    • Manage Site Settings
    • Manage Landing Pages & Favorites
  • Blocks
    • Common Properties
    • Layout
      • Accordion
      • Box & Data Repeater Box
      • Card & Content Card
      • Field & Fieldset
      • Layout Grid
      • Menu
      • Scroll Box
      • Stacked Layout Horizontal & Vertical
      • Tabs
      • Templated List
      • Toolbar
    • Basic
      • Calendar
      • Check Box
      • Color Selector
      • Data Grid
      • Date Selector
      • Dropdown Grid
      • Embedded Page
      • File Library
      • File Uploader
      • Html Editor
      • Image
      • Indicator
      • List
      • Lookup
      • Number Selector
      • Radio Buttons
      • Range Slider
      • Select Box
      • Switch
      • Tags
      • Text
      • Text Area
      • Textbox
      • Tree Grid
      • Tree List
    • Device Input
      • Location Capture
      • Visual Media Capture
    • AI
      • Azure Copilot
      • ChatGPT Copilot
    • Actions
      • Box Hyperlink
      • Button
      • Data Operations
      • Hyperlink
    • Recommendations
      • Alert Action
      • Alert Analytics
      • Alert Discussion
      • Alert Event Data
      • Alert Form
      • Alert List
      • Alert Timeline
      • Alert Triage
      • Alert Survey
      • Recommendation Chart
    • Visualizations
      • Autodesk Forge
      • Azure Digital Twin Hierarchy
      • Bar Gauge
      • Chart
      • Circular Gauge
      • D3 Visualization
      • Esri Map
      • Image Map
      • Linear Gauge
      • Live Feed
      • Map
      • Pie Chart
      • Pivot Grid
      • Polar Chart
      • Power BI
      • Sparkline
      • Time Series Analysis
      • Tree Map
      • Unity
      • Unity (Legacy)
    • Advanced
      • Metablock
    • Widgets
  • Administration
    • Administrative Accounts
    • Language
    • Companies
      • Register a Company
      • Manage a Company
      • Manage Company Subscriptions
      • Manage License
    • Subscriptions
      • Manage User Access
      • Setup Auto Approval/Default Subscriptions
      • Request and Apply a License
    • Users
      • Invite a User
      • Register an Account
      • Profile
      • Change Password
      • Reset Password
      • Delete a User
      • Change Business Role
  • Installation
    • Overview
    • 1. Preparation
    • 2. Install XMPro
      • Azure
      • AWS
      • On-Premise
    • 3. Complete Installation
      • Configure Auto Scale (Optional)
      • Configure Health Checks (Optional)
      • Configure Logging (Optional)
      • Configure SSO (Optional)
        • SSO - Azure AD
        • SSO - ADFS
      • Create Base Company
      • Install Stream Host
        • Windows x64
        • Azure Web Job
        • Ubuntu 20.04 x64
        • Docker
      • Install Agents & Connectors
  • Release Notes
    • v4.4.18
    • v4.4.17
    • v4.4.16
    • v4.4.15
    • v4.4.14
    • v4.4.13
    • v4.4.12
    • v4.4.11
    • v4.4.10
    • v4.4.9
    • v4.4.8
    • v4.4.7
    • v4.4.6
    • v4.4.5
    • v4.4.4
    • v4.4.3
    • v4.4.2
    • v4.4.1
    • v4.4.0
    • Archived
      • v4.3.12
      • v4.3.11
      • v4.3.10
      • v4.3.9
      • v4.3.8
      • v4.3.7
      • v4.3.6
      • v4.3.5
      • v4.3.4
      • v4.3.3
      • v4.3.2
      • v4.3.1
      • v4.3.0
        • v4.2.3
        • v4.2.2
        • v4.2.1
      • v4.2.0
      • v4.1.13
      • v4.1.0
      • v4.0.0
Powered by GitBook
On this page
  • Part 3 – AI at the Core: LLMs and Data Pipelines for Industrial Multi-Agent Generative Systems
  • Part 3 – AI at the Core: LLMs and Data Pipelines for Industrial Multi-Agent Generative Systems
  • Part 3 – AI at the Core: LLMs and Data Pipelines for Industrial Multi-Agent Generative Systems
  • Understanding the Uniqueness of Industrial Processes
  • What is MAGS?
  • From Coding to Commanding: Leveraging LLMs for Enhanced Decision-Making in Industrial Applications
  • Why Data Pipelines are Superior to Traditional Procedural Approaches in MAGS

Was this helpful?

Export as PDF
  1. Resources
  2. FAQs
  3. External Content
  4. Blogs
  5. 2024

Part 3 – AI at the Core: LLMs and Data Pipelines for Industrial Multi-Agent Generative Systems

PreviousUnderstanding the Difference Between XMPro AI Assistant and AI AdvisorNextMAGS: The Killer App for Generative AI in Industrial Applications

Last updated 6 days ago

Was this helpful?

, ,

Part 3 – AI at the Core: LLMs and Data Pipelines for Industrial Multi-Agent Generative Systems

Posted on by

Part 3 – AI at the Core: LLMs and Data Pipelines for Industrial Multi-Agent Generative Systems

Part 3 – AI at the Core: LLMs and Data Pipelines for Industrial Multi-Agent Generative Systems

Pieter Van Schalkwyk

CEO at XMPRO, Author – Building Industrial Digital Twins, DTC Ambassador, Co-chair for AI Joint Work Group at Digital Twin Consortium

Utilizing technology to foster focus instead of allowing it to become a distraction or creating additive sickness. This means avoiding constant alerts from dashboards and warnings from analytics that divert attention away from meaningful work – Mike Carroll

Through parts 3 and 4, I explain XMPro MAGS using an Industrial Predictive Maintenance Agent Team and will explore the anatomy of this Reliability and Root Cause Agent as an example.

Figure 1 – XMPro MAGS Reliability and Root Cause Agent in a Data Pipeline

The following key topics will guide the discussion “Chain of Thought” in this paper:

  • Understanding Industrial Processes: What distinguishes industrial processes from others, and how do the agents within these systems operate differently?

  • Employing LLM-based Agents: How can Large Language Models (LLMs) be utilized within industrial processes?

  • Data Pipelines in Industrial Systems: Why are data pipelines crucial for assembling industrial MAGS?

In part 4 “Pioneering Progress: Real-World Applications of Multi-Agent Generative Systems” I will continue and expand on:

  • Anatomy of MAGS: What are the structural and functional components of MAGS?

  • Examples of XMPro MAGS Applications: What are some real-world applications of XMPro MAGS in industrial contexts?


In my career, I’ve worked with both industrial and non-industrial operational business processes. Here is why I think industrial processes are different and why their agent teams are different.

Understanding the Uniqueness of Industrial Processes

Operational processes in industrial settings markedly differ from typical business workflows found in non-industrial environments. The operational impact of industrial processes often results in safety issues or significant loss, and decisions are made on real-time operations intelligence through SCADA, IoT, Digital Twins, and other operational technology applications.

Traditional business processes are mostly deterministic, designed to follow a predetermined workflow that ensures consistency and efficiency. Such processes execute tasks based on “If-then-else” logic, where every step is predefined and structured to enable consistent performance, even by the least skilled personnel. This model is great for Robotic Process Automation (RPA), where tasks are clearly defined, and outcomes are predictable.

In the image below, I asked Claude to create a Mermaid diagram of the agent process that typical GenAI Agent frameworks like CrewAI, Autogen, and LangChain follow:

Figure 2 – Typical Task-based LLM Agent

These GenAI Agent frameworks are tailored to support defined workflow tasks by generating content, curating, and reviewing within these deterministic processes. These frameworks enhance efficiency by automating routine content generation and curation processes governed by set rules and procedures. Creating the Mermaid diagram is an example of a typical deterministic content generation task that I can assign to one of these agents to improve my efficiency.

“The Decision is the Task” for Many Industrial Processes

In contrast, operational business processes in industrial applications rely on dynamic and complex data inputs. They generally use real-time operational and sensor data that frequently changes, requiring integrating this data with recent contextual information. This information is synthesized and presented to Subject Matter Experts (SMEs) such as Operations Managers, Reliability Engineers, or Quality Supervisors.

These subject matter experts use detailed data, combined with their skills and available documentation, to make informed decisions. It is most often presented in a dashboard to provide the information and context. Their decision-making further involves causal analysis to determine why (Knowing What Causes What, and When – more on this in the future) before taking any action.

We describe these as Emergent Processes, where the next best steps emerge based on the SME’s ability to observe, reflect, plan, and determine actionable options, making these processes fundamentally adaptive and responsive to the changing environment.

MAGS is an example of a GenAI Agent framework that mimics these SME actions. It supports emergent, non-deterministic processes that must handle far more complex and dynamic decision-making than its deterministic counterparts.

MAGS’ primary focus is to use advanced reasoning processes to recommend actions, and in fully automated industrial environments, potentially execute these actions directly. In such scenarios, the SME assumes a critical oversight role as the ‘Human On The Loop’, ensuring that the automated actions align with strategic objectives and operational safety standards.

This shift emphasizes the framework’s ability to operate independently while still under expert governance, enhancing efficiency without compromising control or quality.

What is MAGS?

Multi Agent Generative Systems (MAGS) are systems that consist of multiple agents, each with a defined role, working together to deconstruct and solve complex problems. These systems mimic the hierarchical structure of enterprises and mirror real-world team dynamics. They can also scale intelligently by adding more agents to handle increased workloads.

Each agent in the system handles modular decision-making stepsthat contribute to the overall problem-solving process in a composable way. The agents can further collaborate and cooperate to achieve defined goals without human intervention.

From Coding to Commanding: Leveraging LLMs for Enhanced Decision-Making in Industrial Applications

A common application for existing GenAI Agents, including platforms like CrewAI, Autogen, and LangChain, is optimizing the software development process. These GenAI agents enhance efficiency by specializing in distinct roles within the project workflow.

Project Manager Agents break projects into manageable subtasks, Coding Agents handle the actual development of software, Testing Agents are responsible for quality assurance, and Documentation Agents create comprehensive user guides. This specialization streamlines the entire development process, from conception to completion, ensuring each phase meets high standards of quality and efficiency.

Here, large language models primarily automate tasks like coding and testing, focusing on individual steps without integrating them into broader decision-making processes.

Conversely, Multi-Agent Generative Systems (MAGS) use LLMs for complex decision-making in industrial settings.

For example, an Expert OEE MAGS, which is explained in more detail later, actively improves manufacturing efficiency by deploying specialized AI agents focusing on availability, performance, quality monitoring, and predictive maintenance. These agents can reason and decide on actions based on their observations, reflections, memory, and planning, as shown in Figure 3. They dynamically optimize operations and adapt to evolving industrial challenges by continuously analyzing real-time data.

In systems like the Expert OEE Optimizer, MAGS monitors equipment performance, analyzes data to uncover patterns, devises optimization strategies, and executes these plans, significantly impacting operational outcomes. The primary value of LLMs in MAGS lies not just in performing tasks, but in driving sophisticated reasoning processes that enhance decision-making and operational efficiency.

This contrast highlights significant differences in LLM utilization:

  • Task vs. Decision-Making Focus: In typical non-industrial GenAI agent applications, LLMs enhance task execution. In industrial settings, they support a cycle of observation, reflection, planning, and action, directly influencing strategic, tactical, and operational decisions.

  • Integration and Impact: Non-industrial LLM GenAI agent integration is often compartmentalized, focusing on specific tasks. Industrial LLMs work holistically, with agents interacting and adapting based on collective insights, boosting system intelligence and responsiveness.

  • Scope of Automation: While non-industrial automation targets reducing manual labor in routine tasks (like me drawing the Mermaid diagram), industrial automation through MAGS aims to optimize entire processes, employing GenAI to create effective decision intelligence.

Another key distinction between traditional GenAI Agent frameworks for task-based processes and Multi-Agent Generative Systems (MAGS) lies in their development and deployment methodologies. Task-oriented agent frameworks typically employ a procedural, coded approach that defines LLM content tasks and sequences.

In contrast, MAGS, particularly those developed with XMPro, favor a visual data pipeline approach. This method enables an agile and flexible composition of work teams and agents, enabling more dynamic interaction and rapid adaptation to changing conditions in the operational environment.

Why Data Pipelines are Superior to Traditional Procedural Approaches in MAGS

Traditional task-based Generative AI (GenAI) agents primarily employ procedural coding to execute predefined tasks in a sequence defined by a controller or “supervisor,” as shown in the Mermaid flow diagram in Figure 2. Some platforms assign the “CEO” role to the orchestration functions. While effective for straightforward content generation and curation tasks, this approach struggles in dynamic and complex decision-making applications typical of industrial settings.

In contrast, data pipelines or data streams offer a transformative approach to building MAGS. Unlike procedural coding, data pipelines utilize a fluid, dynamic method that allows for continuous data integration. This integration ensures decisions are based on the most up-to-date and relevant information, which is critical in fast-paced industrial environments. The flexibility of data pipelines allows for modular changes with minimal disruption to the overall system, facilitating easier adaptation to evolving requirements and conditions.

The real power of using data pipelines in MAGS centers on their ability to effectively employ LLMs for complex reasoning and decision-making. Rather than just automating tasks, LLMs in these pipelines can observe data and patterns, reflect, learn, plan, act, and collaborate within a governance framework or “Rules of Engagement”.

This capability significantly improves decision-making by allowing the integration of diverse data sources and tools to contextualize decision data. It facilitates the incorporation of physics-based, statistical, and traditional AI models, anchoring MAGS in practical and empirical reality. This approach ensures that decisions are data-driven and realistically applicable, grounded in physics and reality, greatly enhancing the system’s reliability and effectiveness.

Furthermore, the use of visual, drag-and-drop interfaces in constructing data pipelines offers substantial advantages in terms of usability and explainability. These interfaces allow users to visually map out and modify data flows, making the system’s operations transparent and comprehensible even to those without deep technical expertise. This visual aspect not only simplifies the monitoring and debugging processes but also helps maintain organizational governance.

By implementing data pipelines, organizations can build MAGS that are not only more efficient and adaptive but also easier to manage and scale. The visual building tools and modular, composable design of data pipelines reduce development time and increase consistency across different applications. The ability to visually represent and explain the decision-making process further enhances the value of MAGS, providing clarity and insight into how decisions are derived, which is invaluable for continuous improvement and alignment with business objectives.

While traditional GenAI agents are constrained by the limitations of procedural coding, data pipelines provide a robust framework for developing advanced MAGS. This framework supports dynamic data integration, sophisticated reasoning capabilities, and a visual, modular approach to system design, making it ideally suited to the complexities and demands of industrial applications.

We are opening limited pilot opportunities for innovative, agile leaders in the industry.

July 14, 2024 – Originallly posted on

In , “From Railroads to AI: The Evolution of Game-Changing Utilities,” I shared my thoughts on why GenAI will become a “utility” like electricity and why we have an opportunity to create “appliances” that will change the way we work and improve the quality of life for everyone.

In , “The Future of Work: Harnessing Generative Agents in Manufacturing,” I explained that skills shortages and gaps, productivity challenges, and the complexity of doing business at the speed of thought is forcing us to look at alternate, novel ways to do work and have a “Technology Enabling Focus” as so accurately commented.

As I was writing Part 3 on the architecture of Multi Agent Generative Systems (MAGS) with examples from applications with XMPro, I realized that I first have to explain “What” the reason is for using agents in the first place, that it is not just writing procedural code with some LLM capabilities baked in. So this became Part 3, with following to show the architecture and “How” to build an XMPro MAGS agent team with some examples. I also decided to add Part 5, which will discuss governance, ethical use, trustworthiness, and some thoughts on managing change in organizations planning to use MAGS with the help of

Inspired by this foundational research at Stanford, , XMPro MAGS enables a collaborative environment where multiple agents share insights, learn from each other, and make more informed and contextually relevant decisions. XMPro MAGS has the ability to observe, reflect, memorize, learn, plan, and act using LLMs to simulate the same behavior human SMEs use to make operational decisions.

Figure 3 – XMPro MAGS Agent Observe, Reflect, Plan, Act

In part 4, “,” I continue and show the anatomy of an XMPro MAGS agent team that is built and deployed on this data pipeline and use two examples to demonstrate the capabilities of these agent applications.

If this excites you and you want to be part of the next evolution in industrial applications, please reach out to me or , VP Strategic Solutions.

Linkedin
Part 1
Part 2
Michael Carroll
Part 4
Zoran Milosevic
Generative Agents: Interactive Simulacra of Human Behavior
Pioneering Progress: Real-World Applications of Multi-Agent Generative Systems
Gavin Green
Articles
Blog
CEO'S Blog
July 24, 2024
Pieter van Schalkwyk