Jun 8 – 10, 2021
Europe/Rome timezone

XSpectra®: an Advanced Technology for Improving Real-Time Inspections on Food Production-Lines

Jun 10, 2021, 2:00 PM
20m
Contribution type Session

Speaker

Daniele Macera (Xnext S.p.A.)

Description

The very big need for an improvement in the efficacy of real-time foreign-bodies and non-conformities inspection on production lines in the food sector has pushed Xnext towards the development of a novel patented inspection technology, called XSpectra®. The technology architecture is modular, each acquisition module being equipped with a 128 CdTe pixels 1D array and covering a linear inspection area of 10 cm. The current signals generated by the pixels are read-out and pre-processed by proprietary full-custom Front-End ASICs, whose output signals are then digitized and processed by a full-custom Multi Channel Analyzer providing for radiation spectrum reconstruction. Spectral data of all the acquisition modules is thus conveyed over a proper network interface towards a back-end processor running advanced Neural Network algorithms performing both spectral image reconstruction and foreign bodies detection. Experimental results have shown the XSpectra® capability to operate with a sensitivity down to about 9 keV of energy at photon-rates up to several millions of photons per second. A line-width of about 5.5 keV FWHM has been measured, at room temperature, on the 59.54 keV line of an Am241 very-low activity isotopic source. The system spectral non-linearity error has been measured to be within ±0.5% in the energy range 25 keV - 100 keV. Real-world industrial test-cases have demonstrated the effective superiority of XSpectra®, with respect to other conventional inspection technologies, in detecting low-density foreign bodies inside food products.

Primary authors

Mr Bruno Garavelli (Xnext S.p.A.) Daniele Macera (Xnext S.p.A.) Dr Martina Sammartini (Xnext S.p.A.) Prof. Giuseppe Bertuccio (Politecnico di Milano) Prof. Giacomo Ghiringhelli (Politecnico di Milano) Prof. Andrea Zappettini (IMEM-CNR)

Presentation materials

There are no materials yet.