Data from various sensors can be used to draw conclusions about a machine or system, and sensor data fusion provides the basis for autonomous vehicles. However, the sensor data collected needs to be pre-processed directly on the spot to prevent the network of downstream devices from overloading. Thanks to Industry 4.0, sensor data fusion is only at the start of its success.
Machines and systems are expensive capital goods. If they fail, not only can this disrupt the production process, it can rapidly become extremely expensive for the company. This is where predictive maintenance comes in, with intelligent machines and systems that monitor themselves. This is achieved with a combination of networked sensors, hardware for analysis, and finally, intelligent algorithms. All this in turn requires appropriate processing power. The same also applies for modern vehicles and, in future, autonomous vehicles—here, too, increasingly large volumes of data is provided by various sensors. In order to prevent the electrical system from collapsing under the data load, the sensor data collected needs to be pre-processed using algorithms. This works in the same way as systems and machine monitoring.
The Sensor forum aims to give an overview of sensor data fusion and what approaches it can be used to pursue. It will also show, however, that algorithms and processing power are required for this. Ultimately, the topic of safety needs to be kept in mind.