Articles

How Generative AI and IoT Are Transforming Industry

Written by SDG Group | Jul 28, 2026 10:32:40 AM

The integration of Generative AI and the Internet of Things (IoT) is reshaping industries, enabling real-time intelligence (RTI), predictive maintenance, and prescriptive insights. However, scaling these technologies across enterprises presents unique challenges, from managing massive datasets to ensuring AI reliability at the edge.

 As Principal IoT Architect at SDG Group and a dual Microsoft MVP in IoT and Real-Time Intelligence, Sander van de Velde shares his expertise on how these innovations are transforming industrial operations. 


 “The biggest change we're seeing today is the integration of Generative AI with IoT, turning real-time data into predictive and prescriptive insights. But doing this at scale—that's where the real challenge lies.” 


What is currently the biggest shift in IoT, and what challenges does the industry still face?

The most transformative shift in IoT is the integration of Generative AI into operational workflows. We've moved beyond basic remote monitoring and are now focused on predictive and prescriptive maintenance, where AI analyzes real-time data to predict failures and optimize processes. However, the industry still struggles to scale these solutions.

Working with massive datasets introduces risks such as AI hallucinations, where models generate inaccurate predictions. In addition, deploying Generative AI at the edge, on local networks with limited computing power, adds another layer of complexity. This requires robust frameworks to ensure reliability, especially in mission-critical environments such as ships and manufacturing facilities.

Which technologies are having the greatest impact on our projects?

Two platforms stand out: Databricks and Microsoft Fabric. These technologies enable seamless real-time data ingestion, replacing traditional batch processing. Customers can now move from static reporting to dynamic, AI-driven insights.

For example, we use operations agents as AI-powered virtual assistants to monitor telemetry and execute predefined actions, such as alerting engineers or adjusting parameters. These agents act as "junior virtual engineers," guided by playbooks to ensure consistency. This reduces manual intervention while maintaining operational control.

Can you share a recent project where these technologies solved a critical challenge?

We recently developed a digital twin for an offshore client that needed real-time visibility into its vessel operations. Previously, the company relied on delayed emails and manual reports. Our solution combined live vessel positioning, equipment telemetry, and environmental data into a single integrated model.

The main challenge was designing a flexible rule engine capable of adapting to unpredictable changes, such as weather conditions or equipment unavailability. By decoupling the rules and leveraging edge AI, we created a system that updates dynamically. The client now has real-time visibility into action lead times, significantly improving project planning and resource allocation.

What is the biggest obstacle organizations face when scaling Generative AI within IoT?

Scaling Generative AI from a proof of concept to an enterprise-wide deployment remains the biggest challenge. Small-scale pilots often succeed because they rely on curated datasets, but real-world applications must process raw, unstructured data at scale. This introduces risks such as AI hallucinations, security vulnerabilities, and performance bottlenecks.

To mitigate these risks, we focus on:

  • Edge optimization: Ensuring AI models run efficiently on local hardware.
  • Improved playbooks and guardrails: Safely guiding AI agents through predefined operational rules.
  • Hybrid data strategies: Combining real-time telemetry with contextual datasets, such as ontologies for asset hierarchies.

How has SDG Group's IoT approach evolved over the past five years?

Initially, our focus was on moving data from devices to the cloud—a technical challenge that required expertise in industrial protocols and cloud integration. Today, we've shifted toward transforming raw telemetry into real-time insights using architectures such as the medallion model. We are now taking the next step toward predictive and prescriptive maintenance, helping clients anticipate failures before they occur.

This evolution reflects SDG Group's commitment to seamlessly bridging the gap between Operational Technology (OT) and Information Technology (IT), enabling industries to operate more intelligently, efficiently, and effectively.

About our expert

Sander van de Velde | Principal IoT Architect, SDG Group Nederland

Sander van de Velde specializes in Azure IoT solutions, delivering real-time insights across a wide range of industries. With more than 30 years of experience, he designs and develops IoT platforms using Microsoft Fabric RTI, Azure IoT Hub, Azure IoT Edge, Azure IoT Operations, and Azure Digital Twins.

A Microsoft Certified Azure IoT expert, Sander has been recognized as a Microsoft MVP in Azure IoT since 2017 and in Real-Time Intelligence since 2024. His passion is bridging the gap between OT engineers and cloud data engineers, with a focus on interoperability, remote maintenance, and creating real-time business value.

 Connect with Sander on LinkedIn.