This study investigates the feasibility of chemosensors as a means of detecting order picking errors (type, quantity, condition) regularly observable in intralogistics systems. Therefore, we applied a metal oxide chemosensor originally designed for indoor air quality monitoring to analyse a set of representative packaging materials that the industry commonly uses. The results show that common packaging materials can be classified with an average accuracy of 62.77 %, whereas detecting different quantities (and omissions) showed moderate reliability rates. Concerning order-picking-type-errors, these posed a considerable challenge, particularly when different packaging materials led to similar sensor responses. Condition errors (damages or contaminations) could not be reliably detected due to the variable and unpredictable chemical signatures involved. A key finding is the sensor’s pronounced sensitivity to changes in ambient air composition, which emerged as a factor limiting classification consistency across measurement sessions. The findings suggest that chemosensors have the potential to contribute to error reduction in order picking, particularly as a complementary inspection layer in small and medium-sized enterprises which need cost-effective, lightweight solutions. However, the technology is not yet mature enough for stand-alone industrial deployment, primarily due to limitations in robustness and detection speed under variable environmental conditions.