Project · BMWE / ZIM feasibility study
DynaMaint
Dynamic AI-driven, cloud-based platform for fault diagnosis of HVAC systems and intelligent maintenance coordination – a feasibility study in the Central Innovation Programme for SMEs (ZIM), supported by the Federal Ministry for Economic Affairs and Energy.
Key facts
The project at a glance
- Project: DynaMaint – Dynamic AI-driven, cloud-based platform for fault diagnosis of HVAC systems and intelligent maintenance coordination
- Programme: Central Innovation Programme for SMEs (ZIM) – feasibility study
- Funding body: Federal Ministry for Economic Affairs and Energy (BMWE)
- Project management agency: VDI/VDE Innovation + Technik GmbH
- Project support: FILINA Innovation + Technik GmbH (subsidiary of VDI/VDE-IT)
- Duration: 1 March 2026 to 28 February 2027 (12 months)
- Beneficiary: Dr. Ali Behravan GreeNovaX, Siegen
Background
From reactive repair to predictive maintenance.
Maintenance of heating, ventilation and air-conditioning (HVAC) systems is still largely reactive: faults often go unnoticed for a long time, maintenance jobs are planned by hand, and qualified technicians are scarce. The result is longer downtime, higher costs and wasted energy. According to studies, automated fault detection and diagnosis (FDD) can save 10 to 40 % of HVAC energy use.
DynaMaint addresses this: the planned cloud platform combines real-time AI-based fault diagnosis with intelligent coordination of maintenance teams. A “maintenance team coordinator” not only detects faults and their likely causes, but also supports technicians in planning the right measures. This closes the loop between AI diagnosis and maintenance practice.
Technology approach
The building blocks of the planned platform.
Bayesian fault diagnosis
Fuzzy Bayesian and dynamic Bayesian networks (FBBN/DBN) detect faults probabilistically and identify their likely causes.
Multimodal data fusion
Multimodal neural networks (e.g. LSTM- and Transformer-based) combine sensor data, equipment states, history and text reports.
AI maintenance coordination
Reinforcement learning and dynamic Bayesian networks support scheduling that minimises downtime and balances workload.
Cloud & standards
Scalable cloud microservices and integration via building automation standards such as BACnet/IP and Modbus.
Objectives
What the feasibility study will clarify.
Technical feasibility
Assessing the technical feasibility of an integrated diagnosis and coordination platform based on the current state of science and technology.
Market potential
Analysing the needs of maintenance companies and facility managers, the competitive landscape and potential partners in the German HVAC market.
R&D roadmap
Identifying the R&D work, resources and partners required as the basis for a subsequent R&D project.
Expected benefits
What DynaMaint stands for.
- Shorter downtime through early fault detection and targeted maintenance jobs
- Lower maintenance costs and more efficient use of scarce technicians
- Energy savings in buildings and industry through faults fixed in time
- Greater competitiveness for small and medium-sized building services companies
DynaMaint builds directly on the research of the owner, Dr.-Ing. Ali Behravan, on Bayesian classifiers for fault diagnosis in HVAC systems (doctorate at the University of Siegen).
Funding
Supported by BMWE and ZIM.
The DynaMaint feasibility study is supported by the Federal Ministry for Economic Affairs and Energy (BMWE) on the basis of a decision by the German Bundestag, within the Central Innovation Programme for SMEs (ZIM). The project management agency is VDI/VDE Innovation + Technik GmbH, supported by its subsidiary FILINA Innovation + Technik GmbH.
DynaMaint – ZIM feasibility study
Duration 1 March 2026 – 28 February 2027
Project management agency: VDI/VDE Innovation + Technik GmbH
Let’s talk about your project.
Tell us briefly what you need – we usually reply within two working days.