VASCREEN

VASCREEN technology will allow the detection of potential threats inside luggage or other goods in checkpoints by the analysis of the vapors around. VASCREEN allows the scan of one item (luggage) in less than 1 min based on a radically novel Multidetector Differential Mobility Analyzer (MDMA), coupled with an automatic air sampling system supported by Deep Learning (DL) recognition algorithms.

This new technology has some advantages over the SoA:

• it does not require the removal of electrical items from the luggage,
• it is much cheaper (<0.1€ per item),
• it allows the detection of chemical, biological and explosive threats, including new threats,
• the result does not depend on the interpretation of one operator (automatic),
• it is not intrusive,
• the FAR (False Alarm Rate) is much lower,
• the DR (Detection Rate) is much higher.

During the 6 first months of the project, MION has been working on the adaptation of its technology for luggage screening. To do this, the optimisation of the sampling conditions have been already done after the analysis of many samples using a simulated luggage. In addition, the increase of sensitivity is under validation. The new design has been already manufactured and the first results indicates a significant increase of sensitivity, but no quantification has been done yet.

MION is also working on the development of the final design while EZAKO is working on the implementation of ML models for the classification of the data. Initially, simulated data were used because the proposed instrument was not ready to generate data, but it is expected to have the first dataset from VASCREEN prototype very soon for the validation of the developed algorithms.

The VASCREEN project has allowed the implementation of the technology developed by MION, based on Explosive Vapor Detection (EVDs), to the suitcases and belongings explosive screening supported by the new ML classification algorithms developed by EZAKO. The final demo has demonstrated the capacity of technology to detect vapors of explosive from real suitcases that contain explosive samples at the level of milligrams. This extraordinary detection capacity has been achieved thanks to the optimal combination of MION vapor detector technology with EZAKO’s powerful classification algorithms.

During these 12 months of work, we have been working on:

• The adaptation and optimization of its technology for luggage screening
• The optimization of the sampling conditions
• The implementation of new hardware ideas to increase the sensitivity of the detection
• The implementation of ML models for the classification of the results to increase the selectivity
• The validation of the algorithms on data from a real environment