Introduction

Today's traffic centers control traffic signals in a relatively low-tech environment. However, this situation is fast changing. The traffic management centers (TMCs) of the near future -- under the Advanced Traffic Management Systems (ATMS) component of the Intelligent Transportation Systems (ITS) program -- will be sophisticated facilities with advanced tools and communication systems, manipulating and managing massive amounts of information.

The advanced TMC will help in the real-time management of traffic, including monitoring and controlling roadway access, responding to and managing incidents, rerouting traffic, and communicating and coordinating with the public and the media. It will perform these functions with advanced ITS technology such as sophisticated sensors; data fusion, information processing, and communications equipment; and technology to automate routine decision making and other activities.

TMC operators will have more and better data with which to work than ever before, and they will have a broader range of options for controlling or influencing traffic flow. They will have to interpret incoming information; determine and implement measures to ensure effective traffic flow; and work with various agencies to detect and remove accidents, stalled vehicles, and other incidents that create congestion. These new capabilities and responsibilities far exceed those of today's TMC operator.

Experience -- gained through the automation of the airplane cockpit and the centralized control room -- shows that introducing advanced tools and systems into a previously manually based work environment poses significant human factors challenges. Early phases in the cockpit's evolution were costly ones in terms of human performance, system-induced human error, and accidents. Since then, a new discipline of human factors engineering has arisen to address the human factors problems associated with data display, system integration, and worker performance, among others.

Perhaps the most important lesson learned in previous low-tech to high-tech system conversions is the importance of applying human factors principles and methods early on in the project. An automated system imposed on human workers will not and cannot succeed. A system designed for human users can and will succeed.

To this end, researchers are studying the human factors interface with the higher technology of ATMS in several studies under a "Human Factors in ATMS Design Evolution" contract that was awarded to Georgia Tech Research Institute (GTRI) in fall 1992. These studies include human factors issues in overall system design and function allocation, user interface design, automation approaches, organization design, and operator's capabilities and limitations.

The ATMS Human Factors Contract

Operators sometimes make mistakes while doing their jobs. The design of jobs and systems, however, can have a strong effect on the frequency and significance of these mistakes. The key human factors rule in designing a complex human-machine system, such as the TMC, is to minimize human and system errors and to optimize operator and system performance.

TMC system designers need answers to a broad range of human factors questions. Answers to some of these questions may come from working experience with human factors issues in existing TMCs or other kinds of comparable advanced control rooms. Developing certain design guidelines, though, will require empirical data. Therefore, GTRI researchers are tackling the human factors aspects of the ATMS TMC design project from two main fronts:

• First, researchers are using a "systems engineering design" approach. Under this approach, the human operator is considered a critical component to be integrated into the required analyses and specifications of the TMC design. Thus, the TMC's functional requirements are analyzed in a way that focuses on the operator. The approach is yielding a description of how functions are allocated to humans and machines, how humans analyze their operator tasks, and how to design the TMC to enhance human performance.
• Second, GTRI researchers are conducting experiments and collecting empirical data using a TMC human factors research simulator. This advanced, rapidly reconfigurable simulator was built using data from the systems engineering design research. Findings from using the simulator will feed back into the design effort. The simulator will be used to address issues such as control and display configuration, workstation design, control room layout, level of automation, staffing and job assignment, workload, and operator training and experience.

The main products of the GTRI project will be a human factors design handbook to guide the designers and users of TMCs and a real-time, interactive TMC simulator. Other products will include documented information on all systems engineering design processes, multipurpose database software for future in-house studies, and recommendations for upgrading the simulator for future testing of human factors issues in more sophisticated ATMS TMCs.

The remainder of this article describes the status of the GTRI project and its deliverables.

Systems Engineering Design

The systems engineering aspect of the GTRI project will produce a detailed series of analyses that define a set of TMC objectives, the functions necessary to meet these objectives, opportunities for full or partial automation of the functions, and operator performance requirements and tasks within the partially automated TMC. System automation should be designed to assist and support the traffic management decisions made by the operator. Poor automation can increase the operator's workload and can induce frustration and faulty decisions.

So far, researchers have completed the first of these analyses. They have identified the following objectives for the TMC:

• Maximize the available capacity of areawide roadway systems. The TMC aims to minimize congestion and delay by distributing the traffic load spatially and temporally.
• Minimize the impact of roadway incidents. The TMC aims to reduce the possibility of incidents occurring and to minimize the delay associated with incidents that do occur.
• Help provide emergency services. Interaction with emergency-service providers may include incident detection, verification, and incident notification.
• Help regulate demand. Information from TMC may motivate drivers to reschedule trips, reroute trips, or take alternative modes of transportation.
• Create and maintain public confidence in the TMC. The TMC must be seen as providing accurate and useful information to the public.

Next, the researchers will conduct a functional requirements analysis based on these objectives. What discrete, specific functions are needed to fulfill each objective? For example, fulfilling the first objective requires the capability to assess current traffic and roadway conditions. To assess current traffic conditions, in turn, implies an ability to determine location, speed, and type for each vehicle on the roadway system or for a representative sample of those vehicles.

Once all functions implied by the objectives have been identified, the next step is to allocate those functions to either humans, machines, or a combination of both. Finally, the researchers will determine how well the operator performs the allocated tasks and will decide what is necessary to optimize operator and system performance.

Human Factors Research Simulator

The ATMS human factors research simulator consists of the following components:

• Video simulation server. The server, a Silicon Graphics Onyx RealityEngine2 system, is a workstation dedicated to generating simulated traffic surveillance camera views. It includes a "splitter" that converts the digital simulated video images into analog video signals. These video images may be displayed either on standard video monitors or on the operator workstation monitors via real-time video display cards.
• Four operator workstations. The workstations are on a Silicon Graphics Indigo2 XZ platform and are equipped with touchscreen 486 PCs that allow operators to simulate such physical console controls as pushbuttons and switches. The workstation monitors provide information to the operators on both traffic conditions and the status of the ATMS infrastructure. Operators use this information to monitor and control the simulated ATMS during experiments.
• Experimenter's workstation. This Silicon Graphics Indigo2 XZ workstation is used by the researchers to set up, control, and monitor a traffic scenario during an experiment and to monitor participant performance.
• Traffic model server. This Silicon Graphics Indigo2 XZ workstation is dedicated to running a real-time traffic model (AUTOS). Traffic flow data, calculated by the model, is distributed to the video simulation server, operator and experimenter workstations, and a large display server via a local area network. Commands executed at the operator's and experimenter's workstations modify parameters of the traffic model and thereby affect the traffic simulation.
• Large display server and associated switching device. The Silicon Graphics Indigo2 XZ large display server receives signals, via the switching device, from the various system monitors -- e.g., the video simulation server monitor and the operator workstation monitors. The display is positioned so that it can be seen by all operators.
• Audio communications network. This network supports communications between operators and simulated outside agencies such as police, emergency dispatch, and other TMCs. The experimenter and other research staff members assume the roles of outside agency personnel in communicating with the operators.

The simulator emulates a wide variety of inputs, outputs, and operator support systems. For example, it emulates inputs from traffic and roadway sensors, closed circuit television monitors, voice communication systems (including cellular phones), probe vehicles, and database services. It also emulates outputs to intersection control devices and algorithms, roadway access devices and control algorithms, variable message signs, traffic bulletin boards, commercial radio and television stations, cable television traffic channels, highway advisory radio channels, and voice communication systems.

Development of Design Guidelines

A comprehensive program of human factors research began in June 1994 to help develop design guidelines for future TMC designers and users. The research issues fall into four basic categories: equipment configuration, human-machine performance, job design, and operator training.

Equipment configuration

his study area includes questions about how individual controls and displays should be configured and how control rooms should be organized and arranged. Some candidate issues for experimentation via the human factors research simulator are:

• What types of information should be displayed for TMC operators on a large-screen display -- e.g., an overall network map, a map of only a critical part of the network, video from closed circuit television monitors of key sites?
• What type of control device -- a joystick, mouse, touchscreen, or keyboard -- is best for controlling a remote camera?

Human-machine performance

This study area addresses questions about operator capabilities and limitations in performing different tasks in TMC, with or without machine assistance. Some candidate issues for experimentation are:

• How many incidents can an operator manage at one time?
• How can operators verify incident reports from different sources?
• What is the impact of false alarms from an incident detection support system or the usefulness of that system to the operator?

Job design

Included in this study are questions about how the various tasks that are required of operators should be combined and assigned to specific staff positions. Some candidate issues for experimentation are:

• How involved should the TMC operator be in supervising field maintenance actions? For example, should an operator who is responsible for traffic-signal timing also supervise maintenance of signals and controllers?
• Should responsibilities for TMC operators be assigned geographically -- each is responsible for a given segment of the city -- or functionally -- some operators monitor the roadway system for incidents while other operators manage traffic on the whole network?
• Should cellular telephone incident reports come straight to TMC operators, or should someone else receive these calls and summarize their content for the operators?

Operator training

Included in this study area are questions about which operator skills and abilities are required to perform assigned tasks and what team training is needed. Some candidate issues for experimentation are:

• Should TMC operators' formal training include cross training on all TMC duties and the procedures used by other public agencies -- e.g., incident-service providers?
• How much detailed knowledge do operators need about the software support systems that they must use in their jobs?

Human Factors ATMS TMC Design Guidelines Handbook

A primary product of this research is a comprehensive design handbook of recommendations and guidelines for the designers and users of the future ATMS TMC. This handbook is being developed in two phases. In the first phase, which was completed last summer, analytic data was obtained through: (1) visits to approximately two dozen TMCs and other operational command and control centers and (2) interviews with TMC experts in the United States, Canada, and Europe. Based on these findings, a draft handbook is being prepared. This draft will be modified and refined based on inputs from the study's second phase. During the second phase, empirical data will be obtained by testing technologies and conducting comprehensive human factors experiments on the ATMS research simulator. The resulting final handbook of design standards and guidelines will be available by December 1995.

Summary

The Federal Highway Administration's human factors research program in ATMS is designed to meet the current and projected needs of the TMC design community. Design guidelines are developed based on user requirements and analytical assessments of existing TMCs and comparable systems. The first edition of the human factors ATMS TMC design handbook will be available before the end of 1995, in time to support current upgrades and designs of TMCs. The human factors TMC research simulator supports the development of empirically based human factors design guidelines. This research test bed will continue to support the development and refinement of ATMS TMC design well into the future.

Nazemeh Sobhi is a highway engineer in the Office of Safety and Traffic Operations Research and Development, Federal Highway Administration. Her expertise is in the human factors aspects of intelligent transportation systems. She received a bachelor's degree in computer science from Radford University in 1987 and a master's degree in transportation engineering from the Virginia Polytechnic Institute and State University in 1989. Currently, she is a doctoral candidate in civil engineering at the University of Maryland.