Solutions for Transport Hubs

What we offer

CrowdVision is a leading automated passenger analytics and insights company. CrowdVision’s passenger analytics software shows what is happening to an entire passenger population, in real time. This enables transport hub operators to act decisively to increase efficiency and profitability whilst improving their passengers’ experience.

CrowdVision’s passenger analytics solution processes live video input from commercial off the shelf cameras. The system detects pedestrian movements automatically using sophisticated artificial intelligence techniques. CrowdVision outputs live data about everything from flows, queues and wait times to processing times, occupancies, and asset utilisation.

Live dashboards reveal actionable insights to improve real-time operations. Accumulated historical data provides reliable evidence for planning and investment decisions.

If you are a transport hub or public transport operator, we can enable you to improve your passengers’ experience as well as your bottom line by providing the data and insights that you need to make better decisions. Contact us today to find out how.

CrowdVision’s passenger analytics solution provides real-time data about asset utilisation, for example ticket machines, information booths, and service desks. The solution can report upon utilisation by staff (where applicable) and customers, as well as a range of associated analytics, for example queue lengths and times, and transaction processing times.

CrowdVision’s data about passengers’ use of retail areas can help a transport hub optimise rental returns, as well as enable it to partner with retailers to drive sales. Using data about the number of passengers flowing into retail spaces, and about the number approaching or entering particular retail outlets, and combining it with point of sale basket data, enables the calculation of an end-to-end consideration and conversion funnel. Understanding this funnel is extremely valuable if retail returns are to be optimised.

The CrowdVision’s passenger analytics solution’s real-time dashboards and alerts provide transport hub operators with unprecedented levels of situational awareness, and the opportunity to manage operational issues by exception, as they emerge. For example, real- time occupancy level alerts can be used to manage safety and security issues effectively. Over the longer term, CrowdVision enables transport hubs to improve the efficiency and optimise the profit impact of assets, also leading to improvements in passenger satisfaction.

Our Technology in action

  • CrowdVision processes live video from commercial off the shelf cameras mounted on the ceiling or on pillars.
  • The sensors have a fisheye lens and broad field of view. This keeps deployment costs down.
  • Because sensors are looking down, the solution provides anonymous, non-personal data.
  • CrowdVision’s machine-learning, artificial intelligence algorithms have been trained to recognise people accurately, and distinguish them from other objects, like bags.
  • Our software analyses pedestrian movements in real-time. It counts people and measures their speed and direction of movement.
  • The solution is accurate even in the densest environments, having been originally developed to analyse crowd density during the pilgrimage at Mecca.
  • As part of the CrowdVision software configuration we stitch images from different cameras together to derive a holistic view of an area of interest.
  • We see everyone in that area. It is a 100% sample (all pedestrians) and we know exactly where everyone is at a point in time.
  • This differentiates CrowdVision from Bluetooth/Wi-Fi technologies, which typically measure only a 20-40% sample (not everyone has a phone turned on, whilst others have more than one), and only know approximately where each phone is.
  • Because it is a 100% sample, we can generate statistically-valid results about median performance, as well as best and worst case performance.
  • As part of the CrowdVision configuration we set up boundaries and cylinders to enable the measurement of more analytics than any of our competitors;
    • Flows, into or out of areas of interest,
    • Queues and wait times, both expected and experienced.
    • Asset utilisation, for example at desks, covering both passengers and staff.
    • Processing times, for example at a desk or kiosk.
    • Occupancies and densities in particular zones.
  • Because the video images we process are high quality, even when sensors are mounted high on ceilings, and because our software has been trained using so much data, our analytics are over 95% accurate. Before finalising a configuration, we validate this using extensive manual comparisons of software measurements to human measurements.
  • Having captured CrowdVision data about all the pedestrians in an area of interest, and analysed it, we output actionable data in a number of forms;
    • Real-time dashboards. Live data to use for operational purposes, right now. CrowdVision dashboards are customised for each client to reflect the layout of the physical environment so that operational users quickly understand what the analytics are telling them.
    • Pre-formatted historical reports. Clients also have the ability to use a portal to define their own reports.
    • API data feeds from CrowdVision into data warehouses and Business Intelligence platforms, often so that the data can be combined with other data to generate even richer insights.
  • The CrowdVision solution also offers an input API capability, so that data from other sensors and systems can be combined with CrowdVision’s video analytics on the same dashboards and reports. For example, a client may wish to add Wi-Fi analytics like repeat customer % and measured dwell time to the analytics reported.
  • CrowdVision’s solution runs on the client’s virtualised infrastructure, in real time – either an on-premise data centre or an off-premise private cloud. The main Virtual Appliances for the service are built on an Ubuntu Linux platform.
  • By default, video feeds are not retained once they have been analysed, to reduce storage requirements.
    • The solution is anonymous and privacy compliant, which brings peace of mind as well as alleviating the need for extensive security and personal data audits.
    • The storage that the solution requires is configured as local storage, or as a SAN.
  • Dashboards and reports are published on the client’s intranet, with password-protected access if required. Data can then be shared with business partners if required.
  • CrowdVision is an enterprise solution with built-in back-up processes, redundancy, and system health monitoring.
  • Its security protocols conform with UK Government best-practice, but security settings can be customised to meet a range of scenarios across global regions. Whether you deploy on-premise or in the cloud, you can have confidence in CrowdVision security.
  • The solution is designed to be easily scalable. Measuring additional analytics in an area with sensor coverage requires only software configuration changes. Should a client wish to expand physical coverage, it would simply add more sensors and expand the virtualised infrastructure appropriately.
  • Want to get in touch? CONTACT US