Rethinking how cement plants operate.

Digitizing a cement plant isn’t one product problem. Planning affects production. Production affects quality. Equipment availability affects both. Materials and plant structures need to be understood consistently across applications.

Yet the suite had individual workstreams and Product Owners. That creates a UX challenge that goes beyond individual interfaces.

The need for a cross-product perspective became increasingly visible. Settings, plant structures, materials, navigation, dashboards and AI capabilities kept coming up in internal discussions: topics that could not be solved meaningfully inside individual product silos.

Eight products, one operation. Each covers its own part of running a plant, but they work from the same plant model, the same materials and the same design system.

I became the UX perspective across the product suite. Instead of being permanently assigned to one product, I worked with the Product Owners of the different HProduce workstreams and kept the design perspective on how they connected.

At the same time, I led a UX team of designers and researchers: coordinating work, aligning priorities and helping resolve design problems across products. By 2024, internal resource planning explicitly identified a dedicated multi-person UX and research setup for HProduce.

Without one designer per product, a large part of the job was to orchestrate and prioritize. Through generative research and monthly check-ins with the product owners, I gathered what needed to be accomplished in the coming quarter, turned it into concrete UX tasks with (soft) deadlines on a planning board, and allocated them across the team.

The job itself changed with it: less time on individual screens, more on understanding how each decision affects the system.

That meant regularly stepping above individual requirements. A good example was HProduce for remote operations. Rather than optimizing one isolated dashboard, I pushed the question of what HProduce needs to become relevant for remote operations, so we could identify a set of improvements to prioritize.

Over time, HProduce moved toward shared foundations. Plant and material structures, for example, became cross-product concerns: by late 2025, HProduce had explicit objectives to harmonize its foundational plant and material structure and to make metadata reusable across the suite.

Underneath the products, a shared design system that started in my team keeps patterns and interactions consistent.

The design system →

How many steps a piece of work needs depends on its scale, the resources available and how clear the intended outcome is. A small change can be simple: user feedback, design, specification, implementation. It gets complex when requirements are out of sync with what users expect, which usually shows up in validation and takes one or more iterations.

So we gave generative research enough time before the actual design work started. In the end, that proved to be the cheaper and faster way.

HProduce today spans production planning and optimization, downtime analysis, root-cause failure analysis, quality and production insights, shift reporting and plant KPIs. Its mission explicitly links the suite to operational efficiency, lower production costs and decarbonization potential.

Some solutions within the ecosystem produced significant operational savings. Others helped standardize activities such as failure analysis and build reusable operational knowledge.

For me, the most important outcome was broader: we moved UX from servicing individual products toward helping shape HProduce as a product ecosystem.

The larger the system becomes, the less useful it is to optimize individual parts in isolation.

My biggest contribution to HProduce wasn’t a single interface. It was being able to zoom out, understand how products, users, data and business decisions influenced each other, and then zoom back in far enough to help teams make concrete decisions.

That’s the kind of design work I want to do more of.