Axpo Holding

Modernising Infrastructure through Open Technology Standards

A certain level of secure web-based technology for applications makes a lot of sense, especially in an era where standards like MCP make it easier for applications and AI tools to talk to the systems where our data lives. Local or cloud versions are both possible, and latency is what makes the difference. For data, the open-source Apache stack — Parquet, Arrow, Iceberg — is a clear path forward: it offers the reliability and speed that make today's products possible. On hardware, ARM's advantage in performance per watt is hard to ignore; AWS reports its Graviton instances use up to 60 per cent less energy than comparable instances for the same performance.

Most engineering applications are not there yet, and I expect the organisations that migrate to web-based applications and modern hardware to have an edge. In OT, the protocols and firewalling standards are deeply entrenched and hard to change, and their defined latency classes are fit for purpose. What is needed is the consolidation of the sheer number of valid protocols, especially for IoT sensors, which are still hard to integrate into existing systems.

Improving the Planning, Delivery and Long-Term Management

It sets a standard for us to communicate seamlessly with digital technology. That alone opens the door to automation in a sector that is genuinely hard to automate: the McKinsey Global Institute ranks construction among the least digitised industries in the world. From automated quality checking of construction plans to informed inventory tracking, the gains are real. The structured data being created opens the door to further technologies, especially GenAI, which, on top of this rich data, can build automation for knowledge workers at speeds that were not realistic before — available only because we invested in better data and structure first.

Then there are digital twins, the technology that expresses our need for data integration. They solve problems of undisclosed or missing data, or data that simply stayed in the analogue world. A digital twin is not a shiny 3D model but a way to interconnect data in a form that, in my experience, both experienced and inexperienced colleagues can read. That simplicity is what will mark the difference, including whether we can still work in a sector with an ageing workforce. The IEA counts 2.4 energy workers within ten years of retirement for every worker under 25 in advanced economies. We must train young people quickly and be attractive to them.

Driving Digital Engineering Initiatives across Complex, Multidisciplinary Teams

I have come up with a list of principles that can be found online. But if I have to select the most relevant now, it will be these. Make room for invention on purpose. I have found that even brilliant minds in our sector resist inventing in their unpaid time.

Your task is to move technology from liability to protected value. Start from reality. Tech is an operational cost, and in regulated utilities, we will aim to reduce it at all costs.

Realize that technology integration is expectation management. Software architecture follows managing user expectations and not vice versa. The true cost and value of technology lie in its long-term maintenance, not its launch. This is hard, especially for leaders who are there to make a change (and thrive for new things and less for the next thing to maintain.

Your task is to move technology from liability to protected value. Start from reality. Tech is an operational cost, and in regulated utilities, we will aim to reduce it at all costs. Knowing this departure point simplifies discussions about technology down the road. I have learned that one needs to aim to be first and know when not to. It makes things practical on Budget and when clearly described, it facilitates the Buyin from teams and colleagues.

Building Power Networks on Better Digital Foundations

The biggest opportunity is the transformation to ISO 19650 information management practices, which many know today as the BIM methodology. These practices push power network builders to work more like the tech sector builds software: a common data environment as the single source of information, a clear pipeline, version control, clarity in roles and responsibilities. To that I add the discipline of testing digitally, where changes cost little, and only then building, where they cost a lot. It is the biggest yet untapped chance to change processes, technology and people in power companies.

The challenges lie in decades of neglected technology development for network operators. We still receive working suites built on concepts that predate the cloud and the modern data structures that overtook every application we hold in our hands today. If we bear in mind that that inflection point was roughly 15-20 years, the picture is quite challenging. This makes it harder to adapt existing tech to interconnected workflows, and change almost always requires a gigantic procurement campaign.

Leading Engineers by Setting the Standard

Think that the customer of your customer is your customer. Understand and aim to solve the actual problem of the end user downstream, not just the perceived problem of the project sponsor.

Pump it up into meritocracy, and lead by example. We are engineers, and that is how we tick. Lead from the front, hold a high bar, and contribute to the team's own best work. Suppose you make it until there, proceed to give structure, purpose and direction and then get out of the way. Establish clear rules and cross-functional alliances, set the strategic goals, and let the team execute. The work is not finished at go-live. That is where it starts.