RHODES to intelligent transportation systems

Abstract

I n t e l l i g e n t T r a n s p o r t a t i o n S y s t e m s Intelligent Transportation System program), there have been implicit promises that the implementation will be soon. You only have to read the various articles on smart roads, smart vehicles, vehicle telematics, and so on to realize that the promises haven't yet been kept. For example, one promise was that, by now, most roads would include infrastructure allowing communication from a central system to most vehicles. In addition, most vehicles would be equipped so that through this communication drivers could obtain route guidance, warnings on lane departures, notification of vehicle malfunctions, and so on.To be sure, researchers have conducted tests to show the feasibility of many of the promises, but full-scale deployment is still a ways away. Transportation agencies and automobile companies have realized that maybe they had promised too much and now have more modest goals and milestones. Even the US Federal Highway Administration's latest initiatives, such as the Intelligent Vehicle Initiative (www.its.dot.gov/ivi/ivi.htm) and Vehicle Infrastructure Integration initiative (www.its.dot.gov/initiatives/initiative9. htm), have modest goals for the next five to 10 years.To help fulfill the promises of ITS, the ATLAS (Advanced Traffic and Logistics Algorithms and Systems-see the sidebar for more information) research center is developing and testing the RHODES (Real-Time Hierarchical Optimized Distributed Effective System) traffic control system. 1 We believe that RHODES will play a major role in the realization of future Advanced Traffic Management Systems, a major component of ITS. The future of traffic managementIt's envisioned that future ATMSs will know every vehicle's location (but not necessarily the identification of the vehicle or its driver, unless the driver has provided this information for extra services). Also, traffic management controls and advisories will ensure that vehicles in the network have the smoothest, safest, and most efficient ride from their origin to their destination.The controls and advisories will include• Traffic signals and the phase timings (A set of active signal lights-for example, north-south green with a red left-turn arrow and east-west red-is a phase; phase timing refers to the sequence of phases and each phase's time and duration.) • Roadside or above-road changeable message signs • Highway advisory radio • Pretrip information through radio, television, and invehicle navigation systems • Incident and road work information through radio and in-vehicle systems • En route route guidance through in-vehicle systems ATMSs will obtain traffic information from these sources:• Inductive-loop detectors below the road surface E ver since the major initiatives of the US, Europe, and Japan in the early 1990s to exploit communication, control, and computer advances to make transportation more efficient, reliable, and safe (as embodied in the US

Store-and-forward based methods for the signal control problem in large-scale congested urban road networks

“…Other elaborated model-based traffic-responsive strategies such as PRODYN (Farges et al, 1983) and RHODES (Mirchandani and Head, 1998;Mirchandani and Wang, 2005) employ dynamic programming while OPAC (Gartner, 1983) employs exhaustive enumeration. Due to the exponential complexity of these solution algorithms, the basic optimization kernel is not real-time feasible for more than one junction.…”

Speed Range Prediction for Traffic Light Control System

“…The goal is to minimize the overall waiting time of cars in a city. An adaptive real-time traffic signal control system referred to as RHODES [2], is a "dynamic network loading" model that captures the slow-varying characteristics of traffic. These characteristics pertain to the network geometry and the typical route selection of travellers.…”

Soilse: A decentralized approach to optimization of fluctuating urban traffic using Reinforcement Learning

“…Besides widely deployed UTC systems such as the Sydney Coordinated Adaptive Traffic System (SCATS) [8], [9] and the Split Cycle Offset Optimization Technique (SCOOT) [10], numerous approaches to UTC have been proposed as computational problem-solving methodologies have evolved. Such approaches mainly use Dynamic Programming (DP) [11], [12], [13], [14], graph theory [15], Petri Nets [16] and game theory [17]. Others simply use heuristic models and rule based/logical programming approaches [18], [19].…”

“…[16] uses hybrid Petri Nets to model the traffic network and a supervisor to coordinate all signalized junctions. Several DP approaches to UTC optimiza tion also exist [26], [11], [12], [14], [13], [27]. A reservation based approach for UTC is presented in [28].…”