AI is moving from dashboards to decisions in manufacturing

For years, enterprises have used Artificial Intelligence (AI) to see better. It helped spot patterns, predict failures, and support decision-making. The essential shift is happening now, as AI begins to sit closer to the actual flow of work, from factory floors and supply chains to engineering teams and shop-floor guidance.

For Dassault Systèmes, this evolution is closely tied to virtual twins, governed data, and human oversight. Azmathullah Mohammed, Senior Director, India Customer Solution Experience, Dassault Systèmes, says AI is becoming more embedded into manufacturing, logistics, and workforce support through the 3DEXPERIENCE platform. But he also argues that enterprises cannot trust AI in high-value operations unless the data underneath is accurate, interoperable, and traceable.

In this conversation with Dataquest, Mohammed explains where AI is becoming more operational, why virtual companions matter, and why data integrity remains the foundation for enterprise trust.

Where are you already seeing AI move from insight and support to real operational action?

At Dassault Systèmes, we are seeing AI shift from insight to more autonomous action through Agentic AI and Generative Virtual Twins.  This is also reflected in our early 2026 rollout of Virtual Companions.

In manufacturing, AI is powering predictive maintenance and closed-loop optimisation in smart factories. It can help forecast failures, adjust process parameters through the Industrial Internet of Things (IIoT), and minimise downtime to improve equipment effectiveness. In modular production lines, AI can simulate reconfigurations inside virtual twins before they are deployed physically. This helps manufacturers get closer to first-time-right outcomes and build more flexible production systems.

In supply chains, our solutions use AI for demand forecasting, inventory optimisation, and logistics planning. Algorithms can autonomously balance supply and demand, and automate scheduling, helping shorten delivery cycles by up to 30%. Generative AI agents are also being used to create intelligent workflow automations, including tasks linked to quality control and resource allocation, with human oversight.

AI companions such as Aura are also helping automate repetitive tasks and provide real-time guidance on the shop floor. This improves consistency while capturing expert knowledge, which is important for addressing skill gaps. In India, our research and development centres are helping pioneer these AI agents for global workflows, and we expect more processes to become agent-led in the coming years.

What is the one risk enterprises must solve before they can truly trust AI in core business workflows?

The one risk enterprises must solve is data integrity and interoperability.

Enterprises cannot truly trust AI in core business workflows until they ensure their data is not just abundant, but also unified, accurate, and actionable across silos.

At Dassault Systèmes, trust is foundational to every innovation. As we bring AI into the heart of the our platform and our solution portfolio, we apply the same rigorous commitment to security, privacy, and quality that underpins all our solutions.

Our AI-based software functionality is designed and operated with transparency and human-centric governance. This enables enterprises to use AI confidently and responsibly, with built-in governance controls.

AI works best when it is powered by a single source of truth. Our platform integrates Internet of Things (IoT) streams, digital twins, and AI models into a closed-loop system, where data from design to operations is validated in real time.

We addresses data integrity in Generative AI-powered experiences through a combination of platform architecture, governance, and science-based AI principles embedded in the  ecosystem.

At the foundation is the 3DEXPERIENCE platform, which unifies Product Lifecycle Management (PLM), design, simulation, and collaboration data into a single governed environment. It acts as a single source of truth across the product lifecycle, ensures traceability, version control, and governance of engineering and business data, and enables cross-functional collaboration with controlled access and auditability.

This helps eliminate fragmented data silos, which are among the biggest risks to integrity in AI systems.

We also integrate Generative AI directly into Intellectual Property Lifecycle Management (IPLM). This means AI operates on trusted enterprise data, including designs, simulations, and virtual twins. It protects intellectual property while enabling reuse and generation, and maintains data lineage and ownership across the lifecycle.

This ensures that generative outputs are based on governed, enterprise-grade data, rather than random external sources.