School of Civil and Environmental Engineering

Ph.D. Thesis Defense Announcement

Modeling and Understanding the Implications of Future Truck Technology Scenarios for

Performance-Based Freight Corridor Planning

By

Denise A. Smith

Advisors:

Dr. Frank Southworth (CEE) and Dr. Adjo Amekudzi-Kennedy (CEE)

Committee Members:

Dr. Ram Pendyala (CEE), Dr. Catherine Ross (COA/CEE), Dr. Michael Meyer (Parsons Brinckerhoff)

Date & Time: Tuesday, August 2, 2016, 1:30 PM

Location: Sustainable Education Building, 122

Autonomous highway vehicles are coming. Many advocates predict that autonomous trucks, in particular, will

be commercially available within the next decade. This includes autonomous and connected multi-vehicle truck

platoons. Unfortunately, this technology is developing more rapidly than the public sector is preparing for it: a

situation exacerbated by the fact that the expected arrival of the platoons is within the current planning horizon of

transportation planning agencies. Thus, there is a need to explore the implications of the technology for planning

purposes, which will require the development of tools to quantify potential costs and benefits. With these needs in

mind, the objectives of this thesis were to (1) develop a simulation modeling and performance measurement tool

which incorporates truck platooning technology, (2) demonstrate how this tool can be applied to the I-85 and I-285

corridor in Georgia, and (3) develop a scenario planning framework that uses the results from the tool to guide policy

development. The modeling tool consists of an iteratively linked, multi-commodity and multi-vehicle class truck trip

distribution and a traffic assignment model, requiring changes to the typical travel demand modeling process to

capture the characteristics of platooning technology. The results from an empirical application of this model were

used to assess the safety-, economic-, congestion-, and emissions-related impacts of platooning technology.

The model allowed for variations in platooning details through a multi-variable sensitivity analysis. This

analysis showed a range of costs and benefits of the technology, with the greatest benefits seen when labor costs were

cut by allowing some of the trucks to be driverless. Allowing the autonomous trucks to operate on a dedicated lane

was found to tremendously reduce travel time and congestion for those trucks. In some scenarios, these congestion

benefits came at the expense of the convenience of other vehicles, while in other scenarios, these vehicles experienced

modest congestion-reduction benefits. The emissions impacts varied; the benefits for fuel consumption and emissions

were as much as 9% at optimal speeds. While these findings are insightful, it is important to note that they are based

on a specific set of assumptions. Changing the assumptions in some cases could significantly change the results.

This research is one of the first efforts to modify a traditional travel demand model to simulate autonomous

truck platoons. One of the key components of this contribution is the use of an origin-user equilibrium traffic

assignment, a relatively new path-based assignment which allows the user to specify vehicle class and origin specific

traffic flows, and assign them to the network simultaneously, and which has yet to be explored in depth with respect

to multiple truck class-based, notably platoon-inclusive, freight movements. Additionally, the research presents a new

application of the Freight Analysis Framework, a widely used freight database within the United States. Given the

uncertainty associated with platooning technology, there are various limitations of this research. As the details of

platooning technology become clearer, tools such as the one developed here can help transportation planners better

incorporate such technological advances into their planning process.