Information dynamics in Human Complex Systems
An increasing number of commuters rely on navigation tools like TomTom and applications on smart devices, like Google Maps, to access real time traffic information, forecasts, guidance and support. Through participatory sensing, this broader distribution of personal smart devices leads to more real time traffic information sources and communication possibilities. This means that today's traffic systems are becoming more complex where traffic/commuters are both the sources and consumers of information. However, despite the proliferation of these devices we know very little about the consequences of their complex dynamics.
Transportation systems are human complex systems. This is due to the multitude of participants concurrently interacting at many levels. In a transportation system, the physical network (for example, roads or cars) co-exists with the informational network that is created by the smart devices. This sort of interaction where the participating entities are both the sources and consumers of information, while new in transportation engineering, has been observed in other fields of human activity. For example, the effect of private and public information on financial markets has been the subject of much analysis. Systems like this are generally termed “Human Complex Systems”. Agent based modelling and simulation is the established tool for studying such complex systems. At the Urban Simulations Group in TUM CREATE we believe that similar techniques of modelling and simulation can be used for understanding and creating steering tools for traffic.
Our research focus is on investigating the effect that real time information propagated into the transportation system can have. We study whether disseminating different amounts of information about possible problems can be detrimental for the overall transportation system performance. For instance, if a driver becomes aware that there is a congestion area, they will try to avoid it and select an alternative route. What if all drivers do this? The effect would be that the congestion would simply be moved to a different road. Thus, the question is can we control the amount or form of the information that is transmitted to commuters at any given time to improve traffic efficiency?
The figure represents the schema of the feedback loop between a Transportation System and an Information Control System. On the left side the transportation system is illustrated, where sensors provide data for the block of data acquisition . These sensors can be fixed such as inductive loop detectors, radars, infra-red and so on or they can be mobile and navigation devices. The Information Control System will further pre-process (normalise, clean or fuse) this raw data obtained from sensors and provide cleaned input to the processing block. Following this, the processing block uses this cleaned data to recreate a model of the physical traffic system that reflects the current state of the transportation system. What navigation devices like TomToms and apps like Google maps do is transmit this system state (information dissemination) to the user in terms of routes on a map (albeit based on much simpler data). However, in presenting the information to the user it is acting as an information control system as indicated in the last block. Through a better understanding of the system and the effect of information, the transportation system can be steered towards a better performance
During the course of my thesis I will explore mores aspects related to the feedback loop between the Transportation System and the Information Control System. In future posts, interesting results related to models of information dissemination and information control strategies that are/ can be used in transportation systems will be discussed.