Over recent months Siemens announced three things that look unrelated: Intelligence Center X launched at Realize LIVE Americas 2026, the Industrial AI Suite became generally available on Industrial Edge, and Mendix was positioned as a core component of Intelligence Center X. Put together, they mark a change in what low-code is for inside a manufacturing business.
Mendix is no longer a standalone tool
On its own site, Mendix describes Intelligence Center X as how a company "builds its agentic enterprise", made of three pieces: giving data context, building machine learning models, and developing applications. Mendix holds the third piece, sitting next to RapidMiner's Graph Studio and AI Studio.
The arrangement matters more than it appears. Low-code used to be sold as a way to build applications quickly: you have a problem, you build an app. In the new structure the application is where people and agents meet to act, while contextualised data and trained models sit in the layer beneath. Low-code becomes the interface and execution layer of something larger, not the starting point.
On the platform side, Mendix 11.12 becomes a long-term support release from June 2026, and the Maia assistant has grown from generating code into planning and refining applications.
The infrastructure underneath changed too
At Hannover Messe on 21 April 2026, Siemens announced that the Industrial AI Suite on Industrial Edge is generally available, together with WinCC Unified. The suite aims to compress the whole AI lifecycle and scale it across sites, targeting familiar problems such as predictive maintenance and visual inspection.
In the same announcement Siemens added security functions certified to IEC 62443-4-2 and extended the Industrial Information Hub to bidirectional data flow. This part gets less attention but matters: AI on a plant floor touches control systems, so industrial-grade security is a precondition rather than an added feature.
The numbers that were quoted
In TechHQ's reporting from Realize LIVE Americas 2026, Siemens cited PepsiCo avoiding more than 90% of potential operational issues, improving throughput by 20% and reducing capital expenditure by 10–15%. Simcenter PhysicsAI was quoted as delivering predictions 4,000 times faster at 98% accuracy.
These figures deserve to be read for what they are: vendor numbers presented at the vendor's own event, from an organisation with resources very unlike most manufacturers. Their value lies in the direction they point to, not in the level of return anyone should plan around.
For enterprises in Vietnam
The lesson is not "buy Intelligence Center X". Most plants in Vietnam are not at that stage, and do not need to be. The lesson is the shape of the answer: a context layer connecting the systems already in place, a model layer for a specific problem, and an application layer where people and agents act, with governance running through all three.
That shape works at a far smaller scale. A company with an ageing ERP, a few disconnected operational systems and a data warehouse nobody opens can still start: pick one process that hurts, connect only the data that process needs, build an application the operators actually use daily. Scaling comes afterwards.
The common trap is doing it backwards, buying the platform first and looking for the problem later. Data stays fragmented, agents have no context to reason over, and the result stops at a dashboard nobody opens.
The last point worth noting is governance. All three announcements put auditability and security at the centre rather than treating them as additions. For state-owned enterprises and regulated industries in Vietnam, that is usually what decides whether a project is approved, not how clever the model is.
Sources
Mendix — What's New with Mendix and Intelligence Center X: https://www.mendix.com/whats-new/
TechHQ — Siemens Realize LIVE 2026: Intelligence Center X and digital twins: https://techhq.com/news/siemens-realize-live-2026-intelligence-center-x-digital-twins/
Siemens Press — Siemens Industrial Edge ecosystem strengthens data and AI integration (21 April 2026): https://press.siemens.com/global/en/pressrelease/siemens-industrial-edge-ecosystem-strengthens-data-and-ai-integration