ProMoS NG - Building Intelligence

Seit mehreren Jahren arbeiten wir intensiv an der Entwicklung von KI-basierten Lösungen im Bereich der Gebäudeautomation. Ziel ist es, Gebäude mit Hilfe moderner Technologien und Automatisierung effizienter, komfortabler und nachhaltiger zu machen. In der Anfangsphase wurde das Projekt sowohl finanziell von InnoSuisse als auch mit technischem Know-how der Hochschule Luzern unterstützt. Diese Partnerschaften haben es uns ermöglicht, innovative Ansätze zu verfolgen und die Lösungen für die Anforderungen der modernen Gebäudeautomation kontinuierlich weiterzuentwickeln.

 

Automatic calculation of limit values

Optimization of system control through machine learning for continuous adjustment of limit values based on real-time and historical data.

Machine learning makes it possible to calculate dynamic threshold values for various systems and processes instead of using static, fixed threshold values. The algorithms learn from historical data and recognize the normal operating conditions of a system. On this basis, they can automatically set optimal limit values based on the current conditions. This means, for example, that the limit values for a temperature controller are not just set once, but are dynamically adjusted to take account of changing operating states and environmental conditions. This leads to more efficient and flexible control of systems and enables more precise monitoring of process parameters.

Automatic detection of anomalies

Use of algorithms to identify unusual patterns or deviations in the system data that indicate potential problems or sources of error.

Machine learning algorithms are able to detect anomalies in the system data in real time. They continuously analyze the data streams and compare them with the learned "normal" behavior patterns. If a value is outside the normal range or unexpected behavior is detected, the system automatically sounds an alarm. This is particularly useful for detecting faults or defects in a system at an early stage, which might not be noticed manually until later. For example, faults in the ventilation, heating or lighting technology can be identified quickly before they lead to major problems or failures. The automatic detection of anomalies reduces the effort required for manual monitoring and increases the efficiency and reliability of the systems.

Creating forecasts

Use of historical data and algorithms to predict future events or trends, such as energy consumption in a building or the performance of systems.

The use of machine learning makes it possible to predict future developments on the basis of historical data. This can be applied to a variety of areas, such as predicting energy consumption in a building. Machine learning models analyze past consumption data and recognize patterns that allow them to predict energy consumption for different time periods or under certain conditions. These predictions can help to optimize energy consumption, reduce costs and increase energy efficiency. They can also be used to plan maintenance measures or adjust operating parameters in order to save energy and manage resources more efficiently.