Intelligent methods, solutions and processes for greater efficiency and resource conservation in production.
We focus on the application of various sensor technologies to help the industry to achieve greater efficiency and resource conservation while maintaining flexibility in production. Optical and acoustic sensors are combined into intelligent sensor systems to enable high-precision measurements and to gain new insights into the observed products and processes from the data streams, with the help of artificial intelligence.
- Quality assurance with intelligent sensor technology
- Monitoring of production plants
- Sorting of waste materials
- Human-machine interaction
Imaging sensors have found their way into industry, everyday life and research since the development of the digital camera.
Especially where precise measurements, fast and comprehensible decisions and the connection to computer systems for controlling processes are required, cameras - often in combination with other sensors such as acoustics, laser scanners or geodetic instruments, but also through intelligent placement of the sensors by means of robots - can provide new insights and initiate rapid action.
Machine vision is based on the ability of humans to visually perceive their environment. The Machine Vision and Applications research group uses a variety of machine vision techniques for applications in industrial inspection and measurement, human-centered image analysis, and robotics. The Intelligent Acoustic Solutions research group develops acoustic testing techniques for materials inspection. These non-contact techniques are ideally suited for measuring surfaces, finding defects on workpieces, and controlling robots.
The visual and acoustic measurement techniques are used in tunnel surveying as well as in planning for wooden stairs, they are used on Mars rovers for navigation and mapping, and they measure the amount of rain in so-called "distrometers".
Automated visual inspection of metal surfaces helps control production processes in steel mills, and workpieces such as parquet wood can be checked for quality and defects in seconds during production. Human labour is not completely replaced in the process, but tasks are efficiently shifted to higher levels, which also increases the competitiveness of primarily domestic industry through the research performance of the Institute DIGITAL.
The ever-increasing demand for quality in steel products requires a high level of quality inspection and seamless process control. Automated visual inspection of metal surfaces helps control production processes in steel mills, and workpieces such as parquet wood can be checked for quality and defects in seconds during production.
The intelligent algorithms developed at the institute analyse audio and vibration signals and recognise special noises and acoustic scenes in real time. Using the latest methods from the field of artificial intelligence, the probability of future events is predicted (predictive modelling) and the efficiency of production processes is increased. The combination of several modalities (e.g. audio and video) in one system enables a very high recognition rate of a wide range of events and is used as an innovative monitoring system in a variety of applications for the optimisation of industrial processes:
- Production/Machine Monitoring of industrial plants to track production processes and detect deviations. Projects: BASF, MAHMAMT
- Predictive maintenance Detection of wear and tear and early alerting of faulty processes. Projects: OMV, VOEST
- Industrial quality assurance monitoring of product quality in the manufacturing process.
"Industrial Internet" technologies enable process optimisation through data analysis of sensor streams generated by Internet of Things (IoT).