EN
The New Era of Threat Detection

ARPIA Artificial Intelligence

ARPIA VMI

AI that sees what the human eye might miss.

The next generation of threat detection is here. ARPIA applies advanced computer vision directly to X-ray imagery, making luggage, checkpoint, and cargo inspection faster, more precise, and fundamentally more secure.

Developed by VMI Security, ARPIA analyzes X-ray scans in fractions of a second. It instantly flags weapons, explosives, lithium batteries, and other critical threats enhancing the human operator’s decision-making power without slowing down operations.

Who is ARPIA?

Even the most advanced X-ray hardware ultimately relies on human interpretation. In high-stakes environments, screeners face severe constraints: a standard body scan review takes mere seconds, fatigue sets in, and clever concealment or complex anatomy can mask dangerous threats.
3 main application types

ARPIA is our artificial intelligence platform for automated threat identification across multiple security screening environments. It operates in real time and can be configured for small, medium, and large volumes, adapting to specific operational needs and different numbers of identifiable items.

Small, Medium, and Large Volumes
01

Small, Medium, and Large Volumes

Automatic threat identification for carry-on bags, parcels, and palletized cargo. ARPIA flags firearms, blades, vapes, explosives, lithium batteries, contraband, and miniaturized electronics in real time.

This solution is available in three specialized configurations:

  • Explosives: Detects explosive materials in a wide range of shapes and models.
  • Standard: Provides a broader set of identification classes for general security screening operations.
  • Airports: An AI solution ready for airport environments worldwide, designed to respect the regulations of the aviation security sector.

Each version varies the inspection approach based on a specific need or the number of items to be identified.

Inspection of People
02

Inspection of People

AI for body scanners that locates weapons, sharp objects, micro cell phones, flexible explosive sheets, and illicit items hidden under clothing, inside body cavities, or ingested.

03Trucks and Containers

Overview

ARPIA Cargo is the cargo-focused AI solution within the ARPIA platform, built for customs, border control, port authorities, and fiscal inspection teams. It combines visual AI for X-ray image analysis with textual AI for document analysis, helping operators interpret complex cargo images, verify declared goods, and detect anomalies in real time.

Container and cargo inspection faces two main problems:

  • Fiscal: What is inside the container? Does the physical load match the declared invoice?
  • Security: Is there anything suspicious about the load or cargo?

Typical inspection operations handle high transportation volumes and images that are difficult to interpret. The consequences include tax evasion, long evaluation and approval times, and the release of illicit or dangerous cargo. ARPIA Cargo delegates as many inspections as possible to artificial intelligence.

ARPIA Cargo uses two complementary AI components:

  • Visual Analysis AI: Analyzes X-ray images to classify cargo, locate objects of interest, and flag suspicious or anomalous regions.
  • Textual Analysis AI: Analyzes invoices and shipping documents to classify declared items.
  • Decision Layer: Compares both analyses and returns a clearance conclusion, warning, or alert to the operator.

The image-analysis AI is completely independent of invoice data. It first infers what is present in the image; only afterward is the result compared with the declared documents.

  • Automatic cargo classification and declared-content verification
  • Detection of suspicious areas and anomalies within loads
  • Estimated net weight of cargo based on similar past loads
  • Empty container identification
  • Object localization for categories of interest
  • Estimated object counting for partially homogeneous loads
  • Tax-evasion risk estimation
  • Similar-image suggestions to help operators confirm results
  • Anomaly heatmaps that highlight regions differing from expected patterns
  • Alert generation for inconsistencies
  • Lithium battery detection for cargo configurations