A fundamental problem of machine learning (ML) is measurement uncertainty and the influence on ML results. Measurement uncertainty, which is critical in hazardous gas detection, is directly addressed in this paper. A previously published toolbox is extended for regression. One of the benefits of this approach is obtaining a better understanding of where the overall system should be improved. This can be achieved by either improving the trained ML model or using a sensor with higher precision.

Synchronization problems within distributed sensor networks are a major challenge in the field of Industry 4.0. In this paper, artificially generated time shifts between sensor data and their influence on remaining useful lifetime prediction of electromechanical cylinders are investigated. It is shown that time shifts within sensor data lead to poor remaining useful lifetime predictions. However, this prediction can be significantly improved using various methods as shown in this contribution.

Applications like air quality, fire detection and detection of explosives require selective and quantitative measurements in an ever-changing background of interfering gases. One main issue hindering the successful implementation of gas sensors in real-world applications is the lack of appropriate calibration procedures for advanced gas sensor systems. This article presents a calibration scheme for gas sensors based on gas profiles with unique randomized gas mixtures.

We present a work on gas sensors that can for example be used for the assessment of indoor air quality. These sensors suffer from deterioration by siloxanes, so we investigated these effects by a distinct operation mode and exposition to this gas that allows us to interpret different reactions on the sensor surface. We found that all processes on the sensor surface are slowed down by this treatment and a self-compensation by the evaluation of oxygen adsorption processes is likely to be found.

To fulfil today's requirements, gas sensors have to become more and more sensitive and selective. In this work, we present a novel method to significantly enhance the effect of gate bias on the response of a SiC field-effect transistor by placing a lithium-doped tungsten oxide film beneath the gate. This enhancement, compared to undoped samples, opens new perspectives for static and transient signal generation, e.g. gate bias-cycled operation, and, thus, increasing sensitivity and selectivity.