ITS Research Program, UW

Publication Title:
IVHS Backbone Design and Demonstration

The management of large, complex systems (e.g., power distribution systems, transportation systems) has been greatly aided by the use of increasingly sophisticated sensors and communication technologies to obtain and process real-time data on system activity. Typically, these data are generated by different types of sensors geographically distributed throughout the system. These various sensors take measurements of different system parameters, each of which provides a partial picture of the situation under examination. Under such a system, data typically move from the sensors, through some applications, to the ultimate end user. These data are then used for multiple purposes, including management and control, billing, end-user information, planning, and emergency response.

A wide range of developments have contributed to innovative Intelligent Transportation Systems (ITS). As the ITS community moves toward major demonstration projects and a national architecture, a unifying conceptual framework must be defined. A successful conceptual framework must not only address ITS communication needs, but must also address a number of complex, higher-level needs. These include the following:

1. the need for interagency and multi-jurisdictional data sharing without disruption of existing operations;

2. the need to support existing investments in ITS technology and system development;

3. the need to effectively manage the expansion of sensor technologies and user applications in a straightforward, principled manner;

4. the need to encourage future innovation by providing an open architecture; and

5. the need for interoperability among local, regional, and national ITS developmental efforts.

Research Approach:

We present a conceptual framework for ITS development that has a modular design and addresses the higher-level needs mentioned in the introduction. Again, to meet higher-level needs, the architecture must share multi-jurisdictional data, support existing ITS technology, expand easily, encourage future innovation, and be interoperable among local, regional, and national ITS developments. Our ITS framework has several major components and is based on a client/server distributed computing model. The components include (1) "instance servers" that bridge data sources to a communication network, (2) "fusion servers" that gather and operate on various data types, and (3) presentation systems for delivering management and traveler information. Each client or server process exists in a distributed computing environment, so that they might all be present on one computer or distributed over a set of computers. This distributed architecture allows for the system to be "scaled" and also permits any process to migrate to any computer. Similarly, the number and type of data sources or display applications can easily be extended.

Conclusions and Recommendations:

The approach described in this paper embodies several specific features, which include (1) an architecture that will deliver real-time data to a large number of consumers while maintaining a mechanism for authorization; (2) a definition of distributed applications that includes an entire set of components arranged in an hierarchical structure; and (3) a clear mechanism for building "value-added" applications that use a generally available data stream. In this paper, we have described in detail a set of paradigms to implement such distributed applications and have discussed several actual implementations in the domain of ITS.

We have presented a conceptual framework for ITS development and have shown how this framework solves numerous high-level problems associated with ITS development. A sample instantiation, the Backbone project, further demonstrates the viability of our unified ITS conceptual framework and shows that such a framework can be implemented at reasonable cost and with a high likelihood of successful operation.