What is real-time video analytics in transportation?

Real-time video analytics is an artificial intelligence technology that transforms video feeds from existing cameras at a transit hub into actionable operational data. Unlike traditional video surveillance, which is passive and reviewed after an incident occurs, AI-powered video analytics operates continuously, automatically detects abnormal situations, and generates instant alerts for operations and security teams.

At a train station, airport, or seaport, this technology can count passengers on platforms, detect critical crowding in a corridor, identify abandoned luggage, or anticipate congestion at security checkpoints. All of this is achieved without deploying new physical sensors, by leveraging the existing camera infrastructure.

Why are transportation hubs still “blind” to their own passenger flows?

Most transportation operators already have a dense network of cameras. However, this video data remains largely untapped: it is used for retrospective analysis, not for real-time management. As a result, operational decisions still rely on the intuition of field agents or on imprecise self-reported data.

This is the paradox that AI-powered video analytics directly resolves: the world’s busiest physical spaces possess a goldmine of data that they fail to interpret. As XXII’s vision puts it, a transportation hub “is never silent: it speaks through its data, its flows, and its rhythms.”

What are some concrete use cases for video analytics in train stations?

1. Passenger traffic analysis and anticipation of peak crowds

The CORE platform measures passenger volume in real time by area (platforms, concourses, retail areas, access points, etc.) and compares this data to historical trends to anticipate peak periods. This allows teams to plan staffing and resources proactively, rather than reactively.

Measured result: up to a 30% reduction in delays among operators using the system, thanks to better anticipation of passenger flows.

2. Queue Management and Congestion Prevention

AI automatically detects the formation of lines at security checkpoints, ticket counters, or platform access points. It generates real-time alerts as soon as a critical threshold is approached, allowing for the opening of additional lanes or the redirection of passenger flows before the situation becomes problematic.

3. Optimization of passenger routes

By mapping passengers’ actual routes through the hub, video analytics identifies “breakpoints”—locations where passengers get lost, stop, or turn back. This data helps improve signage, space planning, and service layout.

Measured result: a twofold improvement in passenger flow optimization at equipped hubs.

4. Detection of abandoned luggage and intrusions

In cargo areas, on platforms, or in sensitive public spaces, the platform automatically flags any luggage left unattended without an identified owner beyond a configurable time limit, or any intrusion into a restricted area. Security teams receive an alert with precise location data and an anonymized contextual image.

5. Vehicle counting and classification

For multimodal hubs (airports, ports, major train stations), AI counts and classifies vehicles in circulation (cars, buses, trucks, etc.) to optimize access routes, drop-off zones, and the flow of ground operations.

How does the CORE platform work in a transportation hub?

XXII offers a three-component architecture:

An AI analytics engine (CORE), deployable in the cloud or on-premises, compatible with 100% of cameras on the market (RTSP streams, H.264/H.265, minimum 480p). No camera replacement is necessary.

A visualization interface (BRAIN), a collaborative dashboard that centralizes real-time KPIs: density by zone, historical foot traffic data, detected events, and multi-site comparisons.

A library of configurable use cases: teams can activate the modules they need (crowd management, abandoned luggage, train passenger counting, etc.) themselves with just a few clicks using a no-code configurator.

The solution is operational in less than 15 days after the project launch, with support from dedicated Customer Success Managers and Data Analysts.

XXII and SNCF: A Deployment at the Heart of the French Rail Network

XXII is supporting the SNCF in analyzing passenger flows at its stations. The CORE platform is deployed across the rail infrastructure to track passenger arrivals and departures in real time, monitor platform density, and manage commercial areas. The data collected—train counts per track, platform density, and passenger flow in transfer areas—enables operations teams to manage operations with unprecedented precision, using the SNCF’s existing cameras without any additional hardware investment.

Is video analytics compliant with the GDPR in the transportation sector?

This is the central question for any transportation operator subject to strict legal obligations. The answer is yes, provided that the technology is designed according to the principles of Privacy by Design.

At XXII, this translates in practice to: the systematic anonymization of images displayed in the interface (no data that could be used to identify a passenger by name is exposed by default), data minimization (only data strictly necessary for analysis is collected), documented GDPR compliance with available DPIA and DPA reports, alignment with ISO 27001/27018 standards, encryption of data at rest and in transit, and the complete absence of personal identification.

What results can a transportation operator expect?

There are many teams involved in a transportation hub: operations management, security teams, operations managers, and data & innovation teams. The CORE platform is designed to simultaneously meet the needs of each of these functions from a single interface.

What Video Analytics Does Not Do (and Why It Matters)

Operational video analytics is not mass surveillance. It is not intended to identify individuals, track their personal movements, or build databases of personal behavioral data. Its scope is strictly operational: understanding how passenger flows form, intensify, and dissipate to help teams intervene at the right time and in the right place.

This distinction is fundamental for transit operators who must justify their technology deployments to regulatory bodies and the general public.

Key Takeaways

Real-time video analysis addresses three fundamental challenges at transportation hubs: anticipating early warning signs of congestion before they escalate into incidents; streamlining routes, reducing wait times, and adjusting resources to actual passenger volume; ensuring safety by alerting security teams in real time, without monitoring individual passengers.

The SNCF and other major transit operators have already taken the plunge. For a transportation operations manager, the question is no longer whether video analytics is relevant—it’s how to integrate it into the existing infrastructure without disrupting operations or incurring additional equipment costs.