Microsoft, SAP add more AI to manufacturing, supply chain
At the Hannover Messe industrial show, Microsoft debuted Fabric AI for OT and IT data, and the copilot template for factory operations; SAP unveiled AI capabilities for manufacturing.
AI capabilities continue to become an integral part of enterprise manufacturing and supply chain platforms, as two enterprise heavyweights unveiled significant new features at Hannover Messe 2024.
At the industrial trade show in Hannover, Germany, Microsoft debuted its Microsoft Fabric AI data platform for manufacturing and copilot template for factory operations on Azure AI. Both products are part of Microsoft Cloud for Manufacturing. Microsoft also added AI functionality to Dynamics 365 Supply Chain Management and made new AI capabilities in Dynamics 365 Field Service generally available. The Fabric AI capabilities are intended to offer plant managers data and analytics to help improve operations, and the copilot template is designed to help industrial workers improve productivity.
SAP introduced AI capabilities that are being added to its manufacturing applications, including SAP Digital Manufacturing, SAP Asset Performance Management and SAP Field Service Management. The new AI capabilities are expected to help manufacturers improve operations and supply chains.
AI to help manage OT and IT data
The manufacturing AI capabilities for Microsoft Fabric, an AI-powered analytics and data management platform that launched in November, and the generative AI copilot template for factory operations are in private preview, with general availability expected later this year, according to Microsoft.
Both AI capabilities are aimed at helping organizations manage, analyze and make use of data from manufacturing systems, said Kathleen Mitford, corporate vice president of global industry marketing at Microsoft. The Microsoft Fabric capabilities are at a systemic level and enable manufacturers to integrate their operational technology (OT) data with IT, as well as drive digitization efforts and improve operations.
For example, if a product has come up with quality problems, a product engineer would want to see what issues were coming up on the production line so that they can be addressed in the next iteration, Mitford said.
"Even when you're creating a new version of a product, you can understand what happened in the past," she said. "That data used to be really hard to get to."
The copilot template for factory operations enables companies to develop generative AI assistants that let frontline factory operators use natural language when asking about issues such as maintenance and receive answers from resources including knowledge discovery, documentation, training, issue resolution and asset management systems, Mitford said.
"[The template is] an accelerator for both our customers and partners to apply generative AI on the factory floor," she said.
The new AI capabilities for Dynamics 365 Supply Chain Management are designed to trace a customer product's lineage throughout the supply chain to prevent or mitigate disruptions, according to the company. Customers will use an interface to query and analyze supply chain data. The private preview for the AI capabilities is coming, according to Microsoft.
New capabilities now generally available for Copilot in Dynamics 365 Field Service are aimed at enabling service managers and technicians to find information, resolve issues, keep customers updated and generate summaries of their work.
SAP adds supply chain-focused AI
SAP also unveiled AI capabilities for manufacturing cloud applications that are intended to help customers use their own real-time data to make better decisions to improve product development and manufacturing efficiency, according to the company.
The new AI capabilities are embedded into SAP applications and include the following:
In SAP Digital Manufacturing, customers can use large amounts of machine data and integrate AI to automate and modernize production processes, such as visualizing inspections.
In SAP Asset Performance Management, AI connects and manages IoT devices and sensor data for remote monitoring to improve maintenance strategies and act on predictions or specific conditions, such as detecting anomalies.
In SAP Field Service Management, AI generates predictive routing using historical and real-time data to plan the most efficient routes for the most appropriate technicians.
Manufacturers are looking for turnkey AI applications
Two factors are driving the rise of AI capabilities in manufacturing systems, said Ray Wang, an analyst and founder of Constellation Research.
One is that there's much more digitization in manufacturing, which makes a lot of data available to fine-tune the AI. The other is an increasing demand to use AI to gain business advantages.
"Manufacturers get how data, automation and AI come together," Wang said. "However, they are looking for more turnkey solutions to solve their need to address margin compression, operational efficiency and AI arbitrage."
Microsoft's latest announcements around Fabric enrich their existing capabilities in this area, further reducing the effort required to work with data collected from across different industrial software applications.
Paul MillerAnalyst, Forrester Research
As manufacturers start to scale digital initiatives beyond pilot, proof-of-concept and single-site projects, they face challenges in integrating data across various OT and IT systems, said Paul Miller, an analyst at Forrester Research.
To meet this need, industrial software vendors are starting to offer cloud-based tools that help companies manage this data across multiple applications, Miller said.
"Microsoft's latest announcements around Fabric enrich their existing capabilities in this area, further reducing the effort required to work with data collected from across different industrial software applications," he said.
Generative AI tools such as Microsoft's copilot template for factory operations offer powerful abilities to query and interact with data from across the enterprise, but need to be used with care, Miller said.
Asking a generative AI tool about the state of machines on the factory floor or the time that a shipment will arrive to the warehouse is very different from asking generative AI to draft a marketing email, he said, as the answers must be factually accurate, data-driven and replicable.
"Advances in Fabric that improve the collection, management and use of accurate and authoritative data look, at first glance, less interesting than the GenAI news," Miller said. "But they have far broader utility, as good, clean and well-managed data makes every application and workflow better, whether GenAI-powered or not."
Jim O'Donnell is a senior news writer for TechTarget Editorial who covers ERP and other enterprise applications.