Spectroscopy combined with artificial intelligence
The measurement principle is based on backscatter spectroscopy in the mid-infrared (MIR). The wavelength ranges from 4 to 12 μm could be covered in principle. A core component of the system is a broad-emitting and spectrally fast-scanning laser module consisting of quantum cascade laser (QCL) and micro-opto-electro-mechanical grating scanner (MOEMS). The high brilliance of the light source and the unique properties of the MOEMS scanner enable infrared spectroscopy at a rate of one kilohertz. Such a laser source (spectral tuning range of about 300 cm-1) is integrated in the developed technology demonstrator. However, several individual MOEMS EC QCLs can also be combined with each other so that a much wider bandwidth can be covered with the same measurement time.
The researchers have now combined the core technological components of the system with artificial intelligence (machine vision and machine learning) to enable content verification in the running (inline) process that is both fast and characterized by high confidence. Within 300 milliseconds, a blister consisting of 12 tablets is scanned by spectrally measuring each individual tablet. The acquired measurement data (MIR backscatter spectra) are evaluated in real time.
The scientific results have now been published in the paper titled “Point of Interest Mid-Infrared Spectroscopy for Inline Pharmaceutical Packaging Quality Control” in the IEEE Sensors Journal: