Passenger boarding and loading
The main reasons for capturing data on passenger boarding are to optimise the use of resources, customer service and fare collection, to enable real-time service planning and for timetable and route optimisation. The collection of passenger boarding data allows for detailed route analysis providing information on use of specific routes or route segments at particular times of year, day of the week and time of day. A variety of technologies are currently available to detect passenger boarding and loading including treadle sensors, infrared sensors, internet protocol cameras and stereoscopic cameras. In terms of public transport these all tend to be located on vehicles but can also be located in terminals at entrances. On-vehicle systems are generally separate units that are integrated with other ITS systems via connection to the on-board computer where data can be used for display on the driver console, on in-vehicle displays or transmitted to RTPI systems or the operations control centre.
Treadle sensors/tread mats are generally located on the ground in vehicle doorways. They measure passenger footfall using a pressure switch located in a floor mat. The system can misinterpret the direction of passengers and passenger numbers but despite this retailers can have an accuracy of 95% under most operating conditions.
Active infrared sensors are the most commonly used form of automatic passenger counting technology. These can achieve an accuracy of approximately 95% and are capable of detecting passenger direction, disregarding irrelevant objects such as luggage, operating across a range of lighting conditions and storing information and transmitting information in real-time via wireless networks or through integration with GPS or roadside beacons. The infrared sensor is generally located above a doorway and operates by continuously scanning the area and detecting the temperature contrast between a person and the surrounding environment. The system has no moving parts and is simple to install. Active infrared systems are far more reliable and accurate than passive infrared sensors. Passive infrared detectors emit a beam that acts as a switch as the person enters the vehicle. This type of system may encounter detection difficulties in crowded environments when people are standing in doorways, it may register an object as a person and cannot detect multiple persons entering a vehicle at the same time through the same entrance.
Internet protocol cameras are used alongside a computer system or embedded device for the purpose of conducting passenger counts. Typically they can achieve an accuracy of 90% but can be inaccurate or misinterpret information when there are light level changes. More recent internet protocol systems have been developed where this type of situation does not present a problem.
Stereoscopic sensors are highly effective and are capable of achieving 97% accuracy. This sensor type creates a three dimensional image of the area to determine passenger numbers. It is capable of differentiating between multiple passengers in congested situation. Count data is recorded and can be used in real-time.
The use of in vehicle data capture technologies to detect passenger boarding and loading can be used for multiple applications. They can be used as a passenger surveillance tool to aid drivers, as an operations management tool for route planning and analysis, as a tool for demand responsive transport and to provide passenger information.
Advantages and cautions
Use of passenger counting technologies provides accurate and detailed information on service performance which enables operators to make more informed decisions when considering service changes. Automatic passenger counters provide accuracy levels of between 80% and 97% or greater depending on the choice of technology and they far outperform manual passenger counts and fare box counts in terms of the capabilities and level of detail that they can provide over a given period of time. These systems can however prevent a substantial cost and this is a key consideration.
Relevant case studies
Not observied in the Case Studies
Relevant examples in Ottowa, Canada; Denver and Portland, USA