Gathering accurate and timely information about product and process quality is a strategic asset for every company in today’s customer-centered markets. Recent developments in management information systems and database technology make it feasible to collect data directly from the shop floor. Huge performance boosts have been attained by incorporating, for example, in-memory databases. However, two fundamental challenges still remain to be mastered when connecting shop floor data sources to management information systems. Sampled signals from machines on the shop floor are – on the one hand – unstructured by nature and – on the other hand – highly redundant due to very high sampling rates. This study examines the application of wavelet analysis applied to the injection moulding of thermoplastic materials. It can be shown that the wavelet transform of injection moulding process signals can remedy both the unstructuredness and the redundancy of data derived from sampled signals.