*Part 2 of a three-part series exploring the evolution of Energy Management development on Cumulocity—from foundational platform capabilities to scalable architectures, Digital Twin Manager, AI-assisted development and vertical accelerators.
In our previous blog, we explored why data contextualization is fundamental to building effective IoT solutions. Telemetry on its own has limited value; it only becomes actionable when it’s connected to the physical assets, operational relationships and business context it represents. For energy management applications, that means understanding not just that an inverter is producing 825 kW, but which solar plant it belongs to, how it relates to other equipment, and how its performance contributes to the overall system.
For many engineering teams, however, contextualizing a single site is only the beginning. The real challenge appears when that first deployment proves successful and the solution needs to scale. Suddenly, one solar plant becomes fifty, then two hundred, then a thousand. While each site has its own characteristics, they all share the same fundamental structures. The question shifts from “How do we model this plant?” to “How do we avoid modelling every plant from scratch?”
That is precisely the challenge Digital Twin Manager (DTM) is designed to solve.
The modelling bottleneck
Distributed solar portfolios are deceptively complex. Behind every site sits a hierarchy of assets—plants, arrays, strings, inverters, combiner boxes, weather stations, meters and grid connection points—each with its own metadata, telemetry, alarms and relationships. Before developers can build optimisation logic or customer applications, they first need to represent this physical infrastructure digitally.
Creating that model for the first deployment is valuable engineering work. It forces teams to understand the assets they are working with and establish a consistent structure. Unfortunately, many projects continue repeating that same work for every new customer or site.
Engineering teams often find themselves rebuilding identical hierarchies, recreating metadata, defining the same telemetry structures and manually establishing parent-child relationships. None of this is particularly difficult, but it is time-consuming, and it delays the work that actually differentiates the solution. Every hour spent recreating asset models is an hour not spent developing forecasting algorithms, energy optimisation strategies or customer-facing functionality.
This is one of the hidden costs of custom energy management projects. The bottleneck is rarely collecting telemetry; it’s repeatedly building the foundations that allow that telemetry to be understood.
From digital twins to reusable models
Digital Twin Manager approaches the problem differently. Instead of treating every deployment as a new modelling exercise, it enables engineering teams to create reusable digital t
Consider a typical utility-scale inverter. Once its digital representation has been defined, that model already contains the information developers need: its telemetry structure, asset properties, operational metadata and relationships to other assets within the hierarchy. Rather than recreating those definitions every time a new solar site is commissioned, teams can simply create another instance of the existing model.
The difference may appear incremental, but at scale it fundamentally changes how solutions are delivered. Whether you’re deploying ten inverters or ten thousand, the engineering effort required to model each additional asset is dramatically reduced because the reusable model already exists.
Instead of repeatedly solving the same problem, engineering teams build once and reuse many times.
Building what actually differentiates your business
One of the recurring debates in industrial software is whether organisations should build everything themselves or purchase off-the-shelf solutions. In reality, most engineering leaders recognise that this is the wrong question.
The more useful question is: where does your engineering expertise create competitive advantage?
Very few organisations differentiate themselves by repeatedly modelling inverter hierarchies or defining the same asset relationships for every project. Those activities are necessary, but they are foundational. The real value lies in the capabilities built on top of those foundations—energy forecasting, demand response, predictive maintenance, virtual power plant optimisation or customer-facing digital services.
This is the thinking behind Cumulocity’s Buy & Build approach. Rather than investing months rebuilding common platform capabilities, engineering teams can buy the reusable foundation and focus their effort on the functionality that makes their solution unique. In simple terms buyi the foundational 80% while directing engineering talent toward the differentiated 20%.
Digital Twin Manager applies exactly the same philosophy to asset modelling. Instead of rebuilding digital representations for every deployment, teams establish reusable models that evolve over time and can be applied consistently across future projects.
Scaling engineering, not engineering effort
Imagine managing a portfolio of several hundred distributed solar sites. A new regulatory requirement introduces an additional performance KPI , or a new optimisation application requires every inverter to expose another operational characteristic. In a manually modelled environment, updating every asset becomes a significant engineering task. With reusable digital twin models, the change is made once at the model level and becomes part of every future deployment.
This doesn’t remove engineering from the process—it removes repetitive engineering. Developers remain free to customise applications, extend models and build new capabilities where they create value. What disappears is the need to repeatedly solve the same modelling problems every time another solar site comes online.
For software engineering leaders, this is ultimately what building faster should mean. Not cutting corners or sacrificing flexibility, but reducing the amount of undifferentiated work that slows projects down. As distributed energy portfolios continue to grow, the organisations that scale most effectively will be those that treat digital models as reusable engineering assets rather than one-off project deliverables.
Digital Twin Manager makes that possible, allowing teams to spend less time recreating foundations and more time delivering the energy management capabilities that customers actually notice.