Dr.-Ing. Bernd Petraus, Founder + CTO Digital Building Industries AG (DBI AG)
“District heating or decentralised? Can waste heat sources be integrated?”
Today, local councils face highly complex decisions on such matters: heat pumps, solar PV, storage and volatile prices are all intertwined. And every heat transition is also an electricity transition: thermal output becomes connected load, and the grid connection becomes the critical path. Such decisions are dynamic, costly and best taken today rather than tomorrow. This requires streamlined processes rather than lengthy iterative cycles, at the end of which a PDF ends up on the shelf that is already out of date on the day it is published.
“That’s what AI does.” This usually refers to a language model such as ChatGPT. It promises a great deal, but does not solve the problem: it provides answers that sound plausible but are not verifiable and cannot be proven to be accurate. A black box that cannot be held to account before a committee or client and fails to build trust. Nevertheless, AI helps when used in the right place and supplemented by more suitable methods: machine learning forecasts demand and fills data gaps; system design belongs in the hands of precise optimisation methods. Transparent, verifiable and demonstrably robust, rather than merely plausibly estimated. Using a neighbourhood as an example, we demonstrate how different options can be compared transparently within the committee and where the method reaches its limits. In this way, the planning bottleneck is transformed into a dynamic energy model and, ultimately, a robust, transparent decision that everyone understands and supports.
