The energy transition is rewriting the DSO playbook. Congestion management, flexibility procurement, and local market design are no longer edge cases. They are core operations. And the DSOs that will navigate this shift most effectively are those that treat model-based decision-making as a discipline worth investing in.
While novel AI methods get a lot of attention now, established approaches based on mathematical models are far from exhausted and could yield substantial gains. Mathematical optimisation has long proven its value in transmission and planning contexts. The same potential exists for distribution but realising it requires more than deploying a model. The DSOs that are making real progress share a common mindset. Here is how they think.
They start with the decision, not the tool. The strongest use cases don’t begin with “how do we apply optimisation?” They begin with a specific, costly, or uncertain decision. Think of flexible connection pricing, regional congestion management, investment prioritisation, and work backwards to the right analytical approach. Anchoring to a real problem creates clarity, stakeholder buy-in, and a much shorter path to impact.
They treat explainability as a feature, not a constraint. In regulated environments, a recommendation that can’t be explained won’t be used, no matter how mathematically sound it is. Leading DSOs build this into the design from day one. Simpler, more transparent models that earn trust consistently outperform sophisticated ones that don’t. Sophistication can grow over time; credibility is harder to rebuild once lost.
They close the last-mile gap. Models that live in isolated Python scripts or specialist software rarely survive contact with the organisation. The best DSOs invest in connecting analytical outputs to the tools planners and operators already use. This integration work is where most of the adoption value is won or lost.
They are honest about their data. Distribution grids are still only partially observable. Rather than treating this as a blocker, mature DSO teams scope their models to what the data can support and build in uncertainty explicitly. This keeps outputs credible and creates a natural roadmap for improving data infrastructure over time.
They bring the right people in early. Planners, operators, and regulatory experts are not just end users; they carry the institutional knowledge that makes models reflect reality. Involving them in model design, not just model review, is one of the clearest differentiators between pilots that scale and those that don’t.
The energy transition will demand more structured, data-driven decision-making from DSOs year on year. Optimisation will be a central part of that future. The DSOs that get there first won’t necessarily have the best algorithms, they’ll have built the organisational habits to use good ones well. If you’d like to see what this looks like in practice, our case studies cover work across grid reconfiguration, investment planning, congestion management, and network reliability.