Manufacturing operations generate vast amounts of data at every step in the process, from the design of products to the performance of assembly lines. It is, therefore, a vast challenge to make sense of this data – turning it into actionable insights that will benefit the business.
Significant advances in two areas of technology – digital twins and AI – in manufacturing is releasing some of the constraints that firms faced when trying to use data effectively.
“Manufacturers are no strangers to using data in everyday processes and procedures, but often the data is constrained to specific systems or processes – either physically or digitally,” says Simon Floyd, Microsoft’s general manager of manufacturing and mobility in the Americas. “This situation often hinders the manufacturer’s ability to use the data across processes or functions where it has the potential to have the largest impact.
“With the rise of generative AI tools, manufacturers are now presented with advanced capabilities that utilise natural language interactivity with the combined knowledge of private and public data sources. This provides an entirely new way of interacting with data and creating content that is easy to use.
“The power of the digital twin as the organisational principle for enterprise data is undeniable. By creating a virtual replica of any product, process or machine, manufacturers can simulate specific outcomes and understand their current or future performance. This enables users to train machine learning models to predict failures in specific types of equipment, assess their future performance, and provide opportunities for optimisation and autonomous decision-making.”
Floyd notes that the main benefit for designers who are creating new products or engineers looking to optimise operations is that they will have the ability to analyse data with context, create prediction-based outcomes and simulate different scenarios to ensure an optimal solution.
“Digital twins can take several forms, and one of the most effective examples we see from leading customers are data-based knowledge graphs,” says Floyd. “These graphs use nodes to represent interactions between different assets or factories, providing a clear and organised way to visualise and understand the relationships and dependencies within a manufacturing system.”
By using knowledge graphs, manufacturers can gain visibility and insights into their products or operations in real time. “This approach allows manufacturers to break down data silos and gain a holistic view which leads to improved efficiency and performance,” says Floyd.
Generative AI tools can help manufacturers gain deeper insights from data via a conversational interface – accelerating new design and production options – thereby improving industrial processes and achieving a faster time to market.
“The use of AI in manufacturing is not a new development,” says Floyd. “For the last 30 years, manufacturers have been harnessing the power of AI, in the form of machine learning and deep learning, to automate processes, improve efficiency or quality and reduce costs.
“These early applications laid the foundation for the more advanced and sophisticated uses that we see today. With the introduction of generative AI, manufacturers now have access to advanced AI technologies that enable more natural interactions between humans and machines, that can speed up digital solution development, enable and skill workers faster, and generate synthetic data to power digital twins.”
One resource that manufacturers can use to accelerate their digital transformation is Microsoft Cloud for Manufacturing. “With Microsoft Azure IoT Operations, Azure Digital Twins and Azure HPC, manufacturers can create data representations of physical products, assets and factories,” says Floyd. “They can simulate different scenarios for product design optimisation, process improvement or factory setup decisions, with greater speed and with reduced data sampling.
“Microsoft generative AI capabilities are already embedded across several products. Microsoft 365, for example, provides manufacturers with powerful AI copilots within the different tools they already leverage for their work, such as Microsoft Excel, Teams or Outlook. But for many manufacturers that have already advanced on their data transformation and want to tackle complex use cases with generative AI, we offer the industry-leading portfolio of AI services with Azure AI, including the capability to build your own AI assistant leveraging Azure Open AI Service large language models.”
Digital twin technology and AI tools also allow manufacturing companies to simulate performance based on environmental parameters at different stages of an asset’s lifecycle. This enables real-time monitoring and control, and delivers the ability to learn and improve systems over time using AI and machine learning.
Bridgestone, one of the world’s largest tyre and rubber manufacturers, and Strabag, a European-based construction services provider, have successfully implemented a combination of digital twin technology and AI tools to optimise their processes and improve efficiency.
“Bridgestone has developed a digital twin system for tyre manufacturing, which allows it to simulate and optimise the manufacturing process,” says Floyd. “This has resulted in improved quality control, reduced waste and increased efficiency. Bridgestone is also accelerating product design with AI-connected digital twins and 3D models that allow engineers to simulate and rapidly prototype concepts based on specific criteria before committing physical resources or materials.
“Strabag has leveraged Microsoft Azure to develop a digital twin of its construction sites, enabling it to simulate and optimise construction processes. It is also using AI tools to analyse data from sensors on its construction equipment, predicting maintenance needs and improving maintenance schedules. This has resulted in reduced downtime for its customers and improved safety on construction sites.
“Strabag’s risk management solution, created using Azure OpenAI Service, has also empowered employees to make data-driven decisions, predicting risk with 80 per cent accuracy after just three months of data.”
Microsoft is working with several partners to accelerate time to market with digital twins and AI tools. AVEVA, a global leader in industrial software, is one. The strategic collaboration between the two firms has expanded to provide customers with cloud-based platforms such as AVEVA Connect and Microsoft Fabric. “These platforms streamline the process of collecting, transforming and unifying data from various sources, including manufacturing processes and supply chains,” says Floyd. “They offer tools for data cleansing, transformation and enrichment, thereby making the data more consistent and useful for downstream applications and preparing it to power AI capabilities.”
Microsoft has also collaborated with Siemens to develop the Siemens Industrial Copilot. This AI-powered assistant is designed to enhance human-machine collaboration in manufacturing by generating, optimising and debugging complex automation code. It also provides valuable guidance for maintenance and engineering tasks. “The integration of Siemens Teamcenter software and Microsoft Teams facilitates virtual collaboration across the product design and manufacturing lifecycle, enabling seamless communication and coordination among team members,” says Floyd. “An early adopter of the Siemens Industrial Copilot is Schaeffler, a leading automotive supplier, demonstrating the practical application and benefits of generative AI in the manufacturing industry.”
These collaborations show how generative AI can empower manufacturers and drive advancements in industrial automation and digital transformation.
Floyd highlights that these technologies can quickly gain answers to the most asked questions in real-time manufacturing operations. “Every day people working in manufacturing are asking for answers to common situations such as the performance of a machine, but they wish to know how it compares to the previous hour, day or month,” he says. “Ordinarily this requires tables of data, analysis in a data tool, or pre-determined data views which then require interpretation. Generative AI can create the answer in natural language because it can traverse data faster than any human and use reasoning to generate a response. It comes with the added benefit of data references so enquiring minds can examine the source data or check its validity.
“Imagine how powerful this can be for the many simple questions that are a raised in a factory every day but usually require access to many different data sources just to establish a body of knowledge, which then still requires analysis to form an answer. AI eliminates the lengthy process entirely by providing real answers in just seconds. In any type of operation, whether it’s in the factory, warehouse, or office, this is truly transformative.”
Partner perspectives
We asked selected analysts and Microsoft partners how Microsoft-powered AI tools and digital twin technology are helping manufacturers to improve industrial processes.
“Blue Yonder’s Luminate Cognitive Platform, which runs on Microsoft Azure, enables users to create the data layer that underpins their digital twin, delivers the AI and machine learning that makes their cognitive digital twin smart, and serves as the software-as-a-service foundation that drives speed and scale of scenario planning and decisioning,” said Mathieu Linder, vice president of product management at Blue Yonder.
“Infosys is leveraging Microsoft technologies to design digital twins aligned to the latest industry design principles. Using Microsoft Cloud for Manufacturing, we are bringing the power of immersive, mixed reality capabilities to deepen and accelerate interactivity across the value network,” said Suman Shrivastav, delivery manager, Americas, Microsoft practice at Infosys.
“Microsoft’s mixed reality business applications, such as Dynamics 365 Guides and Remote Assist for HoloLens 2, are revolutionising the manufacturing sector by improving product design, operations and time to market. Moreover, the synergy between internet of things (IoT), digital twins and mixed reality technologies further enhances manufacturing efficiency,” said Christian Segurado, manager of system integration for Microsoft at Mazars.
“Rockwell Automation is collaborating with Microsoft to integrate generative AI in our product suite, notably in the new cloud-native system design product, FactoryTalk Design Studio (FTDS). FTDS provides breakthrough efficiency to engineers, changing the way automation systems are built and delivered,” said Rahul Patel, vice president of software engineering at Rockwell Automation.
“RSM US is utilising Microsoft-powered AI tools, IoT devices and digital twin technology to help manufacturers improve industrial processes and design new products. By implementing predictive maintenance schedules, RSM US ensures optimal uptime for production lines and reduces quality problems due to machine error,” said Nick Bierbrodt, director of business applications consulting at RSM US.
“Sight Machine has integrated its Manufacturing Data Platform with Microsoft’s Azure OpenAI Service to introduce Factory Copilot, a natural language user interface that offers an intuitive, ‘ask the expert’ experience for all users, regardless of their data proficiency,” said Kurt DeMaagd, chief AI officer and co-founder at Sight Machine.
Read more from these partners in the Winter 2023 issue of Technology Record. To get future issues delivered directly to your inbox, sign up for a free subscription.