Rail Security

Innovation Science's Rapid Passenger Tracking software is a mass-transit crime investigation tool.

With its origins invented in response to the 2007 London Underground Bombings, the solution can quickly determine passengers that are of interest to any mass-transit crime investigation. The patented solution is suitable for any complex, highly-interconnected passenger rail network that has a modern ticketing system, and where passengers have the freedom to vary their route during their journey (without revalidating their ticket). This includes major networks such as the London Underground, Hong Kong MTR, Seoul MRT, and many others.

The Rapid Passenger Tracking software minimises analysis time for major incident investigations, and offers practical tools to assist in day-to-day crime prevention.

Solution Synopsis

Although more than a million passengers may be travelling on many of the world’s largest mass-transit rail systems at any given point in time, relatively few of those people are ever relevant to an individual incident that occurs on a rail network. But how do you know which passengers are of interest?

The Rapid Passenger Tracking software efficiently determines exactly which passengers could possibly have been witnesses, casualties and perpetrators of any crime within the ticketed area of a rail network.

Rapid Passenger Tracking Example Screen Captures

The software system delivers timely information to the first-response incident team to answer questions such as:

  • Who may have been present at a particular station, or on a given train during the lead up to an incident?
  • When and where did passengers enter the network who later ended up as casualties?
  • Who may have been involved in the incident that is still on the network?
  • Where and when did passengers who may have been involved in the incident, exit the network?

Further investigations can then be performed using the software's rapid analysis capabilities to answer questions such as:

  • Is it possible for a given passenger to have visited a particular station during a nominated time period?
  • How long could a given passenger have been present at a particular station during their journey?
  • Is it possible for a given passenger to have been in contact with a specified train or a specific other passenger during their journey?
  • Which route did a passenger take during their journey?
  • Can multiple incidents be linked back to a group of passengers that were physically together somewhere else on the network?

Minor crime can also be addressed via the use of the solution's correlation capabilities to detect repeat offenders and help video surveillance personnel focus resources on likely suspects.

Solving the Unsolvable

Tracking millions of people through a complex, highly-interconnected rail network has largely been considered an insurmountable problem — until now. The Rapid Passenger Tracking solution uses unique, patented algorithms and
an entirely new approach to deliver a practical tool for rapidly determining and tracking who is of relevance to any incident that occurs on a subway rail network.

Unlike pure facial recognition solutions, the Rapid Passenger Tracking software predominantly uses ticket turnstile transactions and knowledge of train motion to rapidly isolate ticket holders of interest. The software then identifies small amounts of surveillance footage that are of relevance to each incident investigation. A passenger search takes no more than a few seconds, and video analysis time can be reduced by up to 99.5% over traditional analysis methods.

Deployment & Integration

Operational deployment of the technology requires timely access to passenger turnstile transactions, and train location information (specifically station arrival and departure times for each train).

Detailed route analysis and the provision of a photograph for each candidate passenger requires access to entry turnstile and platform CCTV footage. While there is no requirement for digital CCTV to be employed, route analysis steps are simplified when swift access to digital video sources is available. The solution can be combined with ticket fraud detection measures at entry turnstiles to create a 'virtual ticket' for passengers entering the network without triggering a normal turnstile transaction.

Passenger Privacy

Passengers remain are anonymous throughout the analysis and at no time does the solution require knowledge of individual passenger names or contact details.

Privacy considerations were integrated into the solution from the outset. The actual identities of passengers are not required for the system to perform a complete incident analysis. CCTV footage is associated with ticket identifiers (or surrogate entry tokens) rather than individuals. Most subway rail networks offer the optional registration of reusable tickets to their owners as part of an anti-theft feature. Should local laws permit, the association of this data to ticket identification information could be performed by security agencies as a subsequent analysis step.

Key Features

  • Designed to support the world's most complex subway rail systems.
  • Respects day-to-day passenger privacy.
  • Identifies the ticket holders of relevance to any incident within seconds of an incident being reported.
  • Employs patented search technology.
  • Rapidly isolates small snippets of CCTV surveillance footage from across the rail network that are relevant to any incident investigation.
  • Pin-points an image of every relevant ticket holder using entry/exit turnstile CCTV footage.
  • Provides efficient tools for resolving the precise route taken by every relevant ticket holder.
  • Reduces the amount of CCTV surveillance footage that needs to be analysed for any incident by up to 99.5%.
  • Correlates incidents to identify suspects of recurring crimes (including passenger property thefts).
  • Practical for investigating both major and minor crimes on subway networks.
  • Search algorithms rigorously tested using peak-hour simulations of one of the world's most complicated and busy rail networks.
  • Easily configured to support any major rail network.
  • Designed with a small interface footprint to minimise integration effort.