Empirical Study of Urban Commercial Vehicle Tour Patterns in Texas Wei Zhou, Jane Lin University of Illinois at Chicago Department of Civil and Materials.

Empirical Study of Urban Commercial

Vehicle Tour Patterns in Texas

Wei Zhou, Jane Lin

University of Illinois at Chicago Department of Civil and Materials Engineering

Motivation

• Urban commercial vehicle movements contribute to congestion, adverse environmental, social impacts and safety issues.

• Innovative urban logistics and policies are needed to counteract these negative effects and improve logistical performance with growing goods flows.

• The understanding of urban goods and commercial vehicle movements is thin because of the lack of freight data (often proprietary) and the fact that freight vehicle movements are part of the complex supply chains and logistics activities.

Objective

• To provide an empirical investigation of urban commercial vehicle movements in five metropolitan regions - San Antonio, Amarillo, Valley, Lubbock and Austin – in Texas.

• To quantify how factors such as land use type, shipment demand, cargo type, loading/unloading cargo weight and travel speed affect the commercial vehicle daily trip chaining strategies.

Data Description

• Dataset: Texas Commercial Vehicle Surveys in five counties of San Antonio, Amarillo, Valley, Lubbock and Austin during 2005 and 2006.

• Drivers or operators of the sampled vehicles completed both a vehicle information form and a daily travel log on an assigned day.

• The vehicle information form contains basic vehicle data like vehicle type, fuel type, odometers, etc.;

• The travel log records all trips the commercial vehicle made and all locations they visited during the study day.

Common Key Variables In The Dataset

o Stop-level attributes:

• Longitude and latitude

• Departure/arrival time at stops

• Loading/unloading cargo type

• Loading/unloading cargo weight

• Activity type

• Land use type

Description of Individual Tours

• Two basic types of urban commercial vehicle tours:

1. Direct tour: A direct tour consists of only one stop/visit to a customer before returning to the base;

2. Peddling tour: A peddling tour consists of multiple stops/visits to customers before returning to the base.

Tour Choice Models

• Multinomial logit (MNL) model is built for Texas data with four alternative tour choices:

- Direct tour;

- Peddling tour with two customer stops;

- Peddling tour with three to five customer stops;

- Peddling tour with more than five customer stops.

Explanatory Variables

• Tour length Avg. length Avg. speed

• Land use type Cargo type Activity type

• Tour travel time Dwell time

• Loading/Unloading cargo weight Net Cargo Drop-off Size

• Semi Truck Single unit truck Light duty truck

• Empty trips

Model Results

• Number of Observations : 950

• Log-Likelihood at Constant : -1143.3422

• Log-Likelihood at Convergence : -582.5896

• Rho-Squared w.r.t Constant : 0.4905

• Adjusted Rho-Squared w.r.t Constant : 0.4750

Model Results (Cont’）

  Peddling tour with two customer stops

Peddling tour with three to five customer stops

Peddling tour with more than five customer stops

Variables β StdDev. β StdDev. β StdDev.

Constant -1.039*** 0.1006 -1.0690*** 0.1004 -1.3894*** 0.1149

Farm Products 0.4796 0.6394 -0.2746 0. 5940    

Food, Health, and Beauty Products         1.9353*** 0.5961

Wood Products     0. 2447 0.8655    

Clay, Concrete, Glass, or Stone 2.6144** 1.1380 2.4504** 1.0255    

Goods Pick-up From a Non-Base Location

1.7789*** 0.3946 1.9655*** 0.4111 1.8051*** 0.5177

Goods Pick-up and Delivery From a Non-Base Location

1.5058** 0.6749 1.9896*** 0.6414 1.6856** 0.7577

Visit to Retail -.9057** 0.3732 -0.5061 0.3147    

Unloading Cargo Weight -.5418*** 0.1313 -0.3401 0.1297    

Semi Truck     -1.2558*** 0. 4118 -3.7849*** 0.6179

Empty Trips -0.5536*** 0.1287 -0.1739*** 0.0561    

Conclusions

o The model outputs in this research has led to the following conclusions:

• Urban commercial vehicle tour strategies tend to be associated with cargo type, travel purpose, travel time, dwell time and tour destination.

o The limitations associated with the limitation of the survey data itself in this study requires further research:

• Other variables such as regional land use, urban sprawl, and road network characteristics should be included.

• Some economic indicators like size, revenue as well as other key logistics information such as logistics costs, time windows, vehicle capacity, and driver work hour should be incorporated.

Thanks!