Sensors for monitoring components and the environment have gained importance. BAM has extensive expertise in the development and implementation of sensors. The project Communicating sensor systems for component and environmental monitoring (KonSens) has developed an intelligent platform of hardware and software that can process data from various sources. It relies on communicating systems, uses algorithms for data fusion, and at the same time enables a high degree of mobility.
Sensor systems are important interfaces in the modern world. The term encompasses different technologies. Almost every application uses its own measuring method. The structure and format of the data provided by the sensors are by no means uniform. Something that is not important for individual measurements can be a problem if different sensors are to be used in a common network.
BAM aims to bring order to this data chaos by integrating all of the sensor data on a common platform. The experts refer to this as data fusion. “When we process the data, we find parallels between the applications that are not connected at first glance,” explains Viktor Feller. With this new platform they not only aim to standardise existing measuring systems. The scientists also want to gather additional information. To achieve this, they are combining multivariable data sets generated by different measurement methods.
Thanks to a modular structure, the architecture of the new platform already takes into account future requirements that some of today’s solutions cannot fulfil. It contains all required interfaces and supports numerous communication channels, so that new sensors can be easily integrated. In addition, the platform also facilitates mobility through a high degree of miniaturisation. This is not always easy to achieve. “Some algorithms that run on a standard PC are not suitable for mobile use because computing capacities might be limited,” explains Feller. This is only one of the problems that must be solved when adapting sensor principles from the lab to real life application.
Multi-sensor system measures pollutants in the air
One important application for sensor systems is the analysis of air pollution levels. The issues often vary. For example, the industry is interested in the purity of indoor air and operating air. Protection against harmful substances that could permanently impair the efficacy of a catalyst is also an important issue. The environmental sector on the other hand is concerned with the air pollution in large cities that is caused by volatile gases such as ammonia, hydrogen sulphide, ozone and volatile organic compounds such as benzene.
There are measuring systems that can determine problematic levels, but the demands on the sensors are constantly growing, particularly with regard to the sensitivity of the measurements. Most measurement devices today can reliably detect a few hundred molecules in a gas mixture with one billion particles. But it should also be able to analyse similar pollutant molecules and clearly distinguish them. The sensors should further provide long-term monitoring. “Today’s market-driven technologies are almost at their limits,” explains Feller. That’s why BAM is working on a complex system for multigas analysis. It combines the existing method of specific gas sensor technology with sensors for volatile organic compounds. The combination of both approaches and the joint evaluation aims to create a sensor system for many different measured values.
Electronic file for the monitoring of structures
BAM is also conducting research on the permanent monitoring of structures that are difficult to access, such as offshore wind turbines. At high sea, sensors are used to detect, among other things, corrosion damage and determine the maintenance requirements. This could reduce the number of expensive on-site inspections.
The measurement systems can often be embedded directly in the concrete of the building substance. BAM’s research team uses RFID sensors for this because, according to Feller, they can be embedded over an extensive time span and do not need any internal power supply. “When using these sensors, we follow the principle of an electronic file in the component,” says Feller.
Many of the small sensors remain in the same place for years. If several hundred sensors are used for the monitoring of buildings, it is not enough for them to only supply measurement data. They should also provide additional information that enables the identification and specification of each individual component. Each sensor thus fills its own electronic file. “This is an important requirement for automated data collection when many sensors are involved in monitoring a system,” Feller explains.
Using a sensor network for complex monitoring however, requires a range of development work. Each measurement method requires its own calibration strategy, which is important for ensuring measurement sensitivity. The data provided by the sensor must be subsequently integrated into a system. Before use, it is necessary to determine how and where the sensor is placed in the component or in the environment so that it can collect relevant data. After all, the focus lies on the expert assessment of measurement data and the development of the algorithms for the fusion of the sensor data. “Ten divisions are working on this project,” explains Feller. “Such interdisciplinary cooperation creates a unique opportunity to bring together BAM’s various expertise.”