Tri-Met's Experience With Automatic Passenger Counter and Automatic Vehicle Location Systems

Abstract

The Tri-County Metropolitan Transportation District of Oregon (Tri-Met) is one of about 30 metropolitan transit agencies that have deployed both automatic vehicle location (AVL) and automatic passenger counter (APC) systems (Casey, 1999). These technologies are important components of the agency’s new automated bus dispatching system (BDS). The AVL and APC systems at Tri-Met recover comprehensive operations and passenger activity data at the bus stop level that is archived for later analysis. The agency has gained a reputation as an industry leader in the areas of data archiving and the application of archived data to performance monitoring and analysis. Prior to the implementation of AVL and APC technology, Tri-Met, like others in the transit industry, relied on costly manual data recovery. During lean budget years in the early 1980s, the agency began using stand-alone APCs instead of ride checkers, and recovered very little operating data beyond what was required for Federal Transit Administration reporting. With its AVL and APC systems, Tri-Met is now automatically collecting and archiving over 500,000 stop and event data records per day. Offline analysis of this data supports a wide array of agency activities. The transformation from scarce to abundant data has contributed to a variety of functional changes that have both enhanced service quality and improved efficiency. This report examines factors related to the development of Tri-Met’s BDS that have contributed to its success in data archiving and analysis. The organization of the report includes a description of the BDS, a review of the types of analyses that have been supported by archived data, a discussion of factors that influenced decisions made on the design of the system, reflections on unrealized expectations and missed opportunities, and concluding observations.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 17 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.