Trucks are vital to urban economies by facilitating the movement of goods and services, but their
presence can significantly affect urban infrastructure, traffic congestion, air quality, and noise pollution.
Effective urban planning and policymaking depend on a comprehensive understanding of truck activity.
However, urban truck data collection remains limited. This study will make the first attempt to address this
urban freight data gap by leveraging drive-by sensing technology, specifically leveraging data from Google

Street View (GSV) cars, to analyze truck distribution in urban areas. We aim to develop an advanced vision-
based deep learning framework that can identify the location vehicles and classify them according to body

configurations, which may indicate the type of goods they carry. By utilizing GSV’s extensive visual dataset
and customized deep learning algorithms, we will focus on Manhattan’s complex road network and varied
traffic conditions for precise truck identification. By understanding urban truck distribution, policymakers
can make informed decisions to improve urban living conditions and transportation efficiency. This
research will provide valuable insights into urban truck distribution, enabling policymakers to make
informed decisions to enhance urban living conditions and transportation efficiency. Ultimately, this study
will highlight the potential of combining GSV data with machine learning to advance urban truck activity
analysis and support smart city initiatives