A self-monitoring and self-diagnosis strategy for semiconductor gas sensor systems

We developed a self-diagnostic strategy for intelligent gas measurement systems based on a single semiconductor gas sensor especially for safety and security applications. Selectivity is improved via temperature cycling and/or electrical impedance spectroscopy (EIS). Comparing the redundant data obtained from both methods or from a separate sensor, i.e. a humidity sensor, allows detection of a system malfunction due to sensor damage, caused i.e. by sensor poisoning. If different results (gas type or alarm situation) are obtained a warning (pre-alarm) is given and a self-diagnosis routine is started to identify the cause, i.e. by using another T-cycle or a special signal evaluation. This self-diagnostic strategy improves the robustness of a sensor system and helps to avoid costly false alarms or downtimes. Finally, the acceptance of semiconductor gas sensor systems for safety and security applications will be enhanced with this strategy.