The real-time traffic adaptive control system (RT-TRACS) represents a new, state-of-the-art system in advanced traffic signal control. It was developed cooperatively by a team of U.S. academic, private, and public researchers under the guidance of FHWA. The system provides a frame- work to run multiple traffic control algorithms, existing ones as well as new adaptive algorithms, as they become available. The optimized poli- cies for adaptive control (OPAC) strategy, which provides a dual capa- bility of individual intersection control as well as coordinated control of intersections in a network, was the first adaptive algorithm implemented within the RT-TRACS framework. The operational features of the OPAC prototype version that was developed for RT-TRACS are pre- sented, and its implementation in the Reston Parkway field research test bed in Northern Virginia is described. The implementation provided valuable insights into the performance of coordinated OPAC under var- ious traffic conditions and site geometry. Observations indicate that the strategy was instrumental in reducing delays and stops, compared with a well-tuned fixed time system that was in place, while maintaining pro- gression along the arterial. Valuable lessons were learned that should lead to improvements in future implementations of adaptive algorithms. In the early 1990s, FHWA set out to advance the state-of-the-art in intelligent transportation systems by initiating “a program of research, development and operational tests designed to combat traffic con- gestion.” One of the key elements of this program was “to develop and field evaluate a real-time, traffic-adaptive signal control sys- tem,” which came to be known by its acronym RT-TRACS (1). The objective of the RT-TRACS project was to develop a system capable of adapting to fluctuating traffic conditions by selecting an optimal control strategy from a suite of real-time traffic signal tim- ing control strategies (2). RT-TRACS serves as a platform for imple- menting a variety of traffic signal control algorithms, including new adaptive algorithms as well as existing signal timing systems. PB Farradyne Inc. headed the RT-TRACS project, and the University of Massachusetts at Lowell was assigned the task of developing a proto- type real-time, traffic-adaptive control logic based on the opti- mized policies for adaptive control (OPAC) algorithm. In parallel, FHWA also sponsored the development of alternative adaptive control algorithms for integration into the RT-TRACS platform. Field research tests to evaluate the performance of the different algo- rithms were also included in the program. The first version of RT-TRACS incorporating the coordinated OPAC real-time adaptive algorithm was implemented in a network of 16 intersections on Reston Parkway in Northern Virginia in the spring of 1998. Type 2070 traffic controllers were employed to con- trol a network of signals under both coordinated and isolated modes of the OPAC adaptive algorithm as well as under time-base coordi- nation (TBC). Integration of these technologies into an operating traffic management and signal control system involved a number of technical challenges and institutional issues. This paper describes the basic elements of the OPAC real-time traf- fic adaptive control algorithm that was developed for RT-TRACS and presents an overview of the issues encountered during integration of OPAC into RT-TRACS. Field implementation issues and initial findings of the Reston Parkway field research test are also discussed. OPAC ADAPTIVE CONTROL ALGORITHM The OPAC strategy is a real-time signal timing optimization algo- rithm, which has been under development by the University of Mass- achusetts at Lowell since the 1980s (3, 4). OPAC is a distributed control strategy featuring a dynamic optimization algorithm that cal- culates signal timings to minimize a performance function of total intersection delay and stops. The algorithm uses a combination of measured and modeled demand to determine, in a distributed man- ner, phase durations at each signal that are constrained only by min- imum and maximum green times and, if running in a coordinated mode, by a virtual cycle length and by an offset that can be updated based on real-time data. Development of this strategy was based on the following principles (5, 6 ): • The strategy must provide better performance than off-line methods. Although this principle may appear to be self-evident, it was not always explicitly incorporated in the development of previous responsive strategies. • The strategy must be truly demand responsive—that is, it must adapt to actual traffic conditions and not be responsive to historical or predicted values that are unreliable and may be far from reality. • The strategy must not be restricted to an arbitrary control period (e.g., 10 or 15 min) but should be capable of providing continuously optimized controls. Effective responsiveness cannot be achieved merely by implementing off-line methods at shorter intervals. • Development of new control concepts that are better suited to the variability in traffic flows is required instead of extrapolation of Optimized Policies for Adaptive Control Strategy in Real-Time Traffic Adaptive Control Systems Implementation and Field Testing Nathan H. Gartner, Farhad J. Pooran, and Christina M. Andrews N. H. Gartner, Department of Civil and Environmental Engineering, University of Massachusetts–Lowell, Lowell, MA 01854. F. J. Pooran and C. M. Andrews, PB Farradyne Inc., 3200 Tower Oaks Boulevard, Rockville, MD 20852. 148 ■ Transportation Research Record 1811 Paper No. 02-3667