SMART-KIT | Smart Expansion Joint for Critical Infrastructures

The aim of the "SMART-KIT" project is to develop a connected expansion joint equipped with sensors, whose condition is monitored and possible problems are reported at an early stage.

Coordination: Jan Hämmelmann M.Sc. (CiP)

Duration: 01.08.2023 – 31.07.2025
Funded by: Federal Ministry of Economic Affairs and Climate Action

Initial situation

Expansion joints are indispensable components in many areas of plant and mechanical engineering, as well as a critical component in thermal power plants, petrochemical plants and district heating plants. They are used for low-stress expansion compensation between vibrating and rigid apparatus elements and plant components, whereby they must be optimally adapted to the specific operating conditions. Temperature, pressure and vibration stresses as well as the physical and chemical properties of the operating media have an influence on the design parameters of an expansion joint. Failure of these components can lead to considerable costs, as this usually involves a total breakdown of the plant. Therefore, it is of great interest to monitor the condition of the expansion joint preventively.


The new development of a expansion joint for, among other things, thermal power plants, the operating state of which is recorded and monitored by sensors. This expansion joint is to be monitored using networked digital condition monitoring systems. Thus, a thermal load profile of the compensator can be recorded via a continuous recording of the temperature at a defined number of measuring points and it is possible to forecast upcoming maintenance interventions.


To develop the condition monitoring system, a new expansion joint is developed that is equipped with sensors. Taking data security into consideration, this sensor data is stored in a database. Different load profiles for the expansion joint will be run on a test bench in Dresden. The work packages of PTW are dealing with using this data to realize system monitoring by using machine learning.

Funding source

Consortium partners

Project sponsors