Responsive PI Systems are a new class of passenger information system that works to optimises network performance by using robotic sensor and digital information technologies to inform, influence and coordinate user behaviour. They’re a step up from Static and Dynamic PI Systems that are a standard part of all public transport networks.
The key difference is the ability of Responsive PI Systems is to enhance travel experience for customers and improved operations and network performance. They do this by introducing a communications feedback loop between public transport customers and operators. The loop functions in real time and enables the behaviour of each to influence behaviour of the other.
User behaviour is monitored through sensor networks that feed information into a central coordination program that anticipates where congestion and disruption to operations may occur. This information is fed to public transport operators enabling them to respond to conditions in various ways aimed at easing congestion and reducing potential delays and disruption.
Responsive PI Systems builds on communication technologies for public transport environments that use familiar digital information technologies like smartphones, computer tablets and apps, combining them with robotic technologies.
The essential architecture of robotic systems form the basis of Responsive PI Systems with components that include:
- Sensing & Perception — what does the world look like? What's out there? How many people are on the system and where are they located?
- Cognition — what does the robot think about the world? What's its plan? What advice should be given to people about the state of the system? What do customers need to know so they can alter their behaviour? What do operators need to do to coordinate and support users?
- Actuation — how does the robot physically enact the plan? What type of information can be sent to customers and operators? Should it be sent via smartphones and computer tablets or an audio feed if they need to keep their hands free? Can lights or other signals embedded in the station infrastructure be used to give people subtle signals that will help coordinate people’s movements and keep the system flowing?
Robotics researchers from the Centre for Autonomous Systems have teamed up with transport planners from the Institute for Sustainable Futures to create a Complex Dwell-time Diagnostics (CDD) system that monitors crowding patterns on heavily congested rail stations like Town Hall and Wynyard on the Sydney Trains network.
In some cases, platform crowding slows train services by extending the amount of time trains need to stay at platforms, known as dwell times, to allow passengers to alight and board. Long dwell times mean less train services.
By monitoring behaviour, operators can take pre-emptive action and provide advice to customers on how best to avoid platform crowding enabling everyone to have a more comfortable journey that operates to schedule.
The CDD is a first step towards development of a wider Responsive PI System that would coordinate people and operators across station precincts, concourses and platform areas. Data off sub-systems like the CDD could be given to customers through smartphone and other digital communication devices, helping them to make decisions that will improve their journey experience, and by extension, those of people around them.
The system uses anonymous passenger behaviour data captured using 3D depth senor technology that is fed into a central coordination program that monitors train dwell times, identifying and diagnosing events that cause any disruption or delays to services.
The CDD is currently being developed and tested on-site at Brisbane Central rail station with support from Queensland Rail. The project receives funding from the Rail Manufacturing Cooperative Research Centre (RMCRC) and is being undertaken in collaboration with Downer Rail.