Classical automation technology is strongly tied to manufacturer-specific hardware and software stacks, meaning that modifications to a plant usually require long planning times and costly shutdowns. This rigid approach hinders the maintenance or updating of hardware and software in plants that are often operated for up to 30 years. Software-Defined Automation (SDA) already demonstrates how the decoupling of control logic and hardware works through containerization. Through Software-Defined Automation, updates can be implemented quickly and without shutdowns. While this transforms the control layer, the sensor technology remains firmly coupled to specific hardware. Analog-to-digital conversion, noise filtering, characteristic curve corrections, and all other calculations run locally on specialized hardware. This tightly binds the sensor technology to this hardware and makes algorithm updates time-consuming. 
Software-Defined Sensing (SDS) decouples these two layers. The AD conversion remains at the sensor (physical measurement), but the mathematics and signal processing (filtering, calibration, data interpretation) can be outsourced. This means that new algorithms can be introduced without having to replace the sensor. At the same time, calculation logic can be updated effortlessly, and different sensors can be flexibly coupled with different software pipelines.

GOALS:

• Conception of a hardware and software stack based on a temperature sensor
• Prototypical implementation of the hardware and software stack in two real laboratory demonstrators
• Evaluation of the stack in a test setting

HELPFUL PRIOR KNOWLEDGE:

• Interest in sensor technology, real-time operating systems, and automation
• Solid programming skills, prior experience with embedded controllers, and enjoyment of experimenting with new solutions
• Lectures IT/AT and MARO