A Gamechanger: How Artificial Intelligence Optimizes Heating Systems

Apr 23, 2025

The energy transition in heating is considered one of the greatest challenges facing the real estate industry. While political debates and legislation—like § 60b of the German Building Energy Act (GEG)—set the framework, the real transformation often happens where no one is looking: in the boiler room.

This is where new technologies prove that digitalization is more than a buzzword—it’s a gamechanger. Because if we want to truly reduce energy consumption and CO₂ emissions, good intentions aren’t enough. What’s needed are precise data, adaptive algorithms, and intelligent control systems.

Why AI Systems Are a Gamechanger for Heating Controls

More than 90% of heating systems in Germany are not optimally configured. Their energy consumption is often based on rough estimates or outdated settings—with no consideration for actual usage patterns, building physics, or weather conditions. Especially in the context of the Building Energy Act and § 60b GEG, the potential to optimize existing systems becomes clear.

Artificial Intelligence (AI) can make a decisive difference:

Heating systems are typically controlled via a heating curve, which determines how much heat is generated at which outdoor temperature. To set this curve properly, technicians would need to visit the site repeatedly in different weather conditions to make adjustments. Since this is not practical, systems are usually configured to always provide sufficient heat—which often results in unnecessary energy use.

AI takes over the role of the technician. It is always "on-site," continuously adapting the heating curve in real time. It also detects when a unit is unoccupied or when a resident switches off the heat during the day while at the office. The result: a digital heating control system that intelligently balances consumption and comfort.

Another major advantage: AI detects configuration errors. For example, incorrectly set domestic hot water temperatures—providing heat far beyond what’s needed—can quickly lead to 10% higher fuel costs.

The Benefits Are Clear

Reduction of energy consumption and CO₂ emissions
Automatic adaptation to real usage scenarios
Support for ESG reporting and EU taxonomy compliance
Lower operating costs—without compromising comfort
Fulfillment of legal obligations under § 60b GEG

New Requirements: § 60b GEG & ESG in Focus

With the amendment of the Building Energy Act (§ 60b GEG), regular heating system inspections have become mandatory—especially for larger multi-family buildings. Housing companies and portfolio managers are now required not only to maintain but also to evaluate and optimize their heating systems for energy efficiency. The law explicitly calls for an energy efficiency review—a task that digital systems can complete far more effectively.

At the same time, ESG reporting requirements are increasing. The ability to collect and document consumption and efficiency data is becoming critical for many funds and real estate portfolios.

The challenge: Manual inspections are time-consuming, expensive, and offer no continuous optimization.

The solution: Digital, AI-powered systems automatically provide relevant data—and optimize system performance in real time.

How an AI-Powered Heating Control System Works

Modern solutions typically consist of three core components:

1. Sensors: Temperature, pressure, heat output, and outdoor temperature are recorded at short intervals and transmitted to the cloud.

2. Data analysis & AI models: In the cloud, specialized algorithms run using a “mixture of experts” approach. These are individual pretrained AI modules or physical models, each focusing on a specific task, such as:

Night setback optimization (often overlooked, but highly effective)
Heating curve adjustment
Domestic hot water control analysis
Solar gains based on building orientation and shading (in development)

Each “expert” is designed to achieve a specific optimization goal. What’s unique: The modules work in a cascading decision-making system. Each checks its own domain and only forwards its recommendation if it is validated. This creates a flexible, modular, and precise heating control system that reflects real-world conditions.

Additional modules can be added to detect seasonal consumption patterns or optimize buffer storage strategies and bivalent operation of heat pumps. This modularity allows the system to be gradually expanded and tailored to individual needs.

3. Control & Feedback: Control values are transmitted directly to the heating system—or checked manually if needed. The logic is modular, transparent, and can be adapted at any time.

Our Approach at Sensaru: Simple, Retrofit-Friendly, Effective

At Sensaru, we design systems that translate this intelligent control approach into practical use for the housing industry. Our mission: to make heating systems smart—without having to replace them.

Our solution at a glance:

ThermaSense & Optimizer: Two intelligent sensors that can be installed in under an hour—no changes to the heating system required.
Cloud & AI modules: Our "mixture of experts" architecture was developed in collaboration with the Institute of Control Engineering at KIT and delivers robust, pretrained models.
All-in-one solution: Includes energy reporting, real-time monitoring, and alerts—perfect for ESG integration.

Our data shows that intelligent heating control can reduce energy use by up to 20%, even in older, unrenovated buildings.

Why Pretrained Models Are Better than Self-Learning AI

Self-learning AI systems are often powerful—but they rarely offer transparency into their decision-making.

Our advantage: Our AI modules are pretrained, meaning we ensure in advance that the models respect defined limits and don’t cause tenant complaints. With self-learning models, there's always a risk of performance degrading unexpectedly.

Our models are trained using curated datasets gathered and validated over defined periods. The result:

Faster deployment, as models don’t require long learning phases in operation
Easier updates, since any module can be replaced with a newer, better version

Checklist: Is Your Building Ready for Smart Heating Control?

If you answer “yes” to the following questions, it’s time to consider an AI-powered heating solution:

Are there heating systems in your portfolio older than 10 years?
Do you lack regular or transparent consumption data?
Is the heating curve undocumented or unclear?
Are ESG or EU taxonomy requirements relevant to your operations?
Are tenants or regulators demanding better energy performance?

Conclusion: The Heat Transition Starts with the Right Data

If you're serious about reducing the CO₂ footprint of your real estate portfolio, intelligent heating control is essential. AI offers a fast, scalable, and cost-effective solution—especially for existing buildings.

Whether for retrofits, ESG strategies, or compliance with laws like § 60b GEG—the combination of digital heating control, smart automation, and modular technology makes the heat transition achievable in the real estate sector.

At Sensaru, we believe: The future of heating is in the cloud—and it starts with a smart sensor in the boiler room.