Automatic Passenger Counting (APC) Systems - An Overview

Automatic passenger counting (APC) refers to technology that automatically records the number of passengers boarding, alighting, and riding in public transport vehicles, replacing manual passenger surveys with continuous, real-time data. Seven main APC technologies are in use today: light barriers on vehicle doors, pressure sensors in seats, train weight measurement, radio signal detection, ticket apps, payment barriers at network entries and exits, and video analytics using security cameras. This article compares their accuracy, cost, and limitations, and outlines the AI-powered video analytics approach used by Isarsoft. For implementation details specific to your fleet, see our deep-dives on trains and buses.

Why passenger counting is necessary

A thorough analysis of traffic flows is needed to ensure that public transport is designed to meet demand. With accurate passenger data, routes and fleets can be adjusted to the real traveling locations. Additionally, in associations with several providers, the association revenues must be distributed fairly among the individual providers. To ensure this, it is common practice to conduct passenger surveys at certain intervals.

Typical questions that need to be solved in this context:

• Number of persons in the vehicle (utilization)
• Fluctuation at each stop
• Ticket type
• Entry, and exit point
• Age of passengers
• Occupation
• Gender
• Use of other means of transport

Since such manual passenger surveys have to be carried out over a long period of time and over a wide area to be able to draw conclusions, they are often accompanied by a large financial outlay. A further complicating factor is that even with a high survey density, individual aspects that do not occur permanently or only locally cannot be recorded. To overcome these problems, there is an effort to automatically and continuously capture as many of the above issues as possible.

Solutions for automatic passenger counting (APC)

Over time, different approaches for APC have developed with different strengths and weaknesses, which we will cover in this article.

A short overview over the different technologies for APC.

Light barriers on vehicle doors

A relatively simple way to measure the number of people in the vehicle is to install a light barrier in each door. Depending on the deviation from the undisturbed state, the system detects whether a person enters or leaves the vehicle. With the help of the fluctuation per stop, the total number of passengers during a trip can be measured.

The weakness of such systems is that the simultaneous entry and exit of several people is often not detected correctly. Bicycles and baby carriages are also a problem, as they are not detected by the light barrier. In addition, there are the additional costs for the sensor technology.

Not to be neglected is that each incorrectly detected entry and exit has a negative effect on the accuracy of the total number of people. Over the duration of a trip, the error can increase steadily.

Pressure sensors in the seats

An alternative way of measuring occupancy is to use pressure sensors in the seats. If a seat is occupied, the sensor inside sends a signal to the associated control unit.

The disadvantage of this method is that only the seat occupancy rate is considered. Especially in rush hours, however, a large part of the passengers are often standing. Furthermore, the costs of such a system are relatively high because each seat must be equipped with a sensor.

Train weight

Each person increases the total weight of the train. Measuring the total weight of the train and subtracting the empty weight gives a rough estimate of the occupancy of the train.

Radio signals

Most people carry electrical devices such as smartphones that emit signals like WLAN, Bluetooth or LTE. By measuring these signals, you can infer the number of people in the vehicle.

However, some people carry multiple electronic devices or none at all, which decreases counting accuracy.

Ticket apps

Apps for purchasing tickets on passengers' smartphones can be used as a source of information for passenger counting. However, in rural areas as well as in tunnels, the internet connection and GPS location can be poor or distorted, making it difficult to measure the occupancy of individual vehicles. As long as analog ticket sales still exist, ticket apps provide only partial information.

Payment barriers at network entries and exits

Payment systems like the one in the picture below can record the number of people on the network. Unique payment card identification numbers can also record the origin and destination of each passenger.

For such an approach, each entrance and exit to the network must be equipped with appropriate payment barriers, which means that the financial outlay is correspondingly high. In addition, it must be ensured that in an emergency, the payment barriers do not stop passengers leaving the buildings. This may even require structural measures.

Another disadvantage of such an approach is that the utilization of individual vehicles is not recorded.

Video analytics

Camera-based video analytics can be applied at three different points in a public transport network: at network entrances and exits, at vehicle doors, and inside the vehicle itself. All positions build on the security camera infrastructure, but each position offers a different tradeoff between speed of implementation and depth of data. Security cameras at network access points capture overall passenger inflow and outflow faster and more easily than payment barriers, though without tracking individual routes. Cameras mounted in vehicle doors, typically positioned centrally and facing downward, use image processing to detect whether each passenger is entering or leaving, while cameras placed inside the vehicle measure total occupancy directly. We cover this approach in detail for trains and buses, including camera placement and setup.

‍

How Isarsoft solves the cost-accuracy tradeoff

Choosing the right passenger counting technology ultimately comes down to a tradeoff most operators already know well: accuracy, cost, and speed of deployment rarely align in traditional sensor systems. Camera-based AI removes that tradeoff because the camera infrastructure already protecting your passengers can now also power your planning, reporting, and revenue-sharing decisions. Without the need for new hardware, investment costs can be kept low. This is the approach Isarsoft has built specifically for public transport operators, and it's already running in vehicles and networks today. See how Isarsoft supports public transport operators.