Open Letter from the CEO

As a young mechanical engineer, I was put in charge of the maintenance function of a large petrochemical plant very early in my career. I was from a new generation of engineers that had the opportunity to work with microcomputers or PCs (many of which we built ourselves and learned to code by trial and error). I was part the first generation of engineers who worked on spreadsheets (VisiCalc and Lotus 123), and we built our first databases on dBase III and dBase IV. It was groundbreaking as I created a full Reliability-Centred Maintenance (RCM) graph for the petrochemical plant. My maintenance planners were the first to use PCs to help us operate the plant at its “Full Potential”. Or so I thought. I was Digital Engineer 1.0

It’s been a long and interesting journey since those early “predictive maintenance” spreadsheets. I’ve witnessed the rise and demise of many methodologies, tools, architectures, and applications that claim to be the holy grail for asset and process performance management. We’ve seen monolithic “enterprise resource planning” solutions that claim you only need a single solution to address all challenges and hundreds of custom applications then followed that all created “islands of automation.”

These “islands of automation” created new challenges in having a Common Operating Picture with a holistic view and coordinated objectives to run the plant at its “Full Potential.” Digital twins go a long way to address that. XMPro has been developing advanced digital twin capabilities and business process technology since 2009 to bring a Tesla-like single-view experience to industrial operations. We built tools to support Digital Engineer 2.0.

We have an event-driven focus, so we look for business events that can blindside operations. These events are either happening, or likely to happen. These “events” present opportunities, and threats or challenges. We have become exceptionally good at enabling real-time digital twins. We are now providing the next generation of Intelligent Digital Twins that are always-on, proactive, goal-seeking, and, most importantly, powered by AI to process, analyze, and recommend the best next actions to run your plant at “Full Potential.” We are empowering Digital Engineer 3.0.

Artificial intelligence (AI) brings a game-changing new capability to support and augment the work of our best engineers and subject matter experts. It even brings the opportunity to have different levels of autonomous operations in our plants and factories and to “operate by wire” in a safe operating envelope. It is the first time since the spreadsheet in the 1980s that I see technology that will change how we work at a foundational level. AI is not a better mouse trap like ERPs and all the other three-letter acronym soup; it is a fundamentally different way of operating, leveraging digital twin capabilities to provide interaction between people, process, and technology.

At XMPro, we embrace the opportunities from traditional Machine Learning (ML), modern Deep Learning, and the latest Large Language Models (LLMs) popularised by ChatGPT. We obsess over how to make these and other emerging technologies useful, scalable, and ultimately sustainable. We understand that AI needs to solve real business problems. For that reason, AI needs to be embedded in existing and new business processes, foster innovation, and augment the work your smartest employees are doing.

XMPro AI: Foundational Capability of XMPro iDTS

I’m proud to introduce XMPro AI, the latest XMPro iDTS (intelligent Digital Twin Suite) capability. XMPro AI empowers real-time, intelligent applications within a digital twins framework, offering:

• Composability: XMPro AI is designed to easily integrate or ‘plug-in’ to existing processes and data streams, leveraging XMPro’s comprehensive AI connector library.
• Innovation and Experimentation: XMPro AI is compatible with XMPro Notebooks, facilitating model development, simulation, and seamless integration with pre-existing libraries of engineering and mathematical models.
• Productivity Enhancement: XMPro AI enhances operational efficiency by integrating domain-specific large language models, such as Azure OpenAI services and ChatGPT, to deliver real-time recommendations and guidance.

All these capabilities are designed and implemented from a digital twin perspective, thereby encapsulating all the advantages associated with a digital twin approach.

Here are some of the key benefits of XMPro’s AI-enabled intelligent digital twins for industrial applications and use cases that provides a single Common Operating Picture, across the organization’s value stream:

• Streamlined Process Enhancement: AI-enabled digital twins simulate various process scenarios and determine the most efficient configurations in real-time. They allow organizations to experiment with different settings and parameters in a virtual environment, identify bottlenecks, and optimize process flows without disrupting actual operations.
• Enhanced Asset Utilization and Predictive Maintenance: Digital twins paired with AI continuously monitor the condition and performance of physical assets. This proactive approach to maintenance identifies potential failures before they happen, ensuring maximum asset uptime and optimal performance.
• Superior Quality Control: AI models trained on digital twin data identify potential quality issues early in the production process. This early detection rectifies defects before products reach the market, leading to improved quality control, cost reduction, and increased customer satisfaction.
• Responsible and Sustainable Governance: Digital twins simulate and evaluate operations’ environmental, social, and governance (ESG) impact. AI analyzes these simulations to provide recommendations for improving ESG compliance and sustainability.
• Minimized Operational Risk: Digital twins model operational risk scenarios accurately, and AI analyzes these models to predict and prevent potential hazards. These mitigations span risks like equipment failures, safety hazards, and cybersecurity threats.
• Safe Environment Assurance: Digital twins help create safer work environments by simulating various safety scenarios and identifying potential hazards. AI analyzes these simulations and suggests improvements, preventing accidents and ensuring the safety of employees.
• Efficient Supply Chain Management: AI-enabled digital twins create end-to-end virtual models of supply chains, tracking products from raw materials to customer delivery. AI uses this model to optimize supply chain operations, reduce lead times, and respond proactively to disruptions.
• Optimized Workforce Empowerment: Digital twins create digital representations of the workplace, helping to optimize workforce management. AI uses this data to predict workforce needs, improve task allocation, and optimize schedules, resulting in increased productivity and employee satisfaction.

XMPro AI is a practical approach that builds on XMPro iDTS core capabilities of integrated, intelligent, interactive, and composable digital twins.

XMPro AI Is Seamlessly Integrated Into Your Business Processes

One of the significant hurdles associated with the surging tide of standalone AI applications is the necessity for business users to navigate numerous user interfaces and applications to exploit these novel capabilities. It harks back to the era when our phones were flooded with apps. Initially, it was exciting that there was “an app for that,” but we soon found ourselves endlessly scrolling through pages to find the right one.

Standalone AI applications present analogous difficulties. They might be packed with impressive features, but their lack of integration with business processes, possible through a comprehensive digital twin approach, detracts from their utility. AI should be seamlessly interwoven into business procedures to ensure consistency, repeatability, and scalability.

With XMPro AI, incorporating AI commences with the data flow in XMPro Data Stream Designer. You can effortlessly drag and drop prebuilt AI and ML libraries, as well as third-party services such as AWS SageMaker or Azure ML, onto the real-time data stream. Our no-code environment allows easy configuration to connect it to historians, sensors, business systems, and contextual data sources. Manage your models in MLOps tools like MLFlow and execute them directly in the data streams