Truck platooning in the U.S. national road network: A system-level modeling approach

Highlights

• Develop a system-level equilibrium model to characterize spontaneous truck platooning.

• Account for relationships among platoon formation time, fuel saving, and road capacity increase.

• Solution method is proposed with a diagonalization approach and bush-based algorithm.

• A new zoning system is constructed by employing spatially constrained multivariate clustering.

• Application to the US suggests about a billion dollar operating cost saving annually.

Abstract

Truck platooning enables a group of trucks to move close together, which helps reduce truck fuel use and increase effective road capacity. In this paper, a system-level equilibrium model is developed to characterize spontaneous truck platooning with coexistence of non-platooning vehicles in a network, by explicitly accounting for the interlocking relationship among platoon formation time, truck fuel saving, and increase in effective road capacity. To equilibrate the relationships, an algorithm is proposed which involves a diagonalization approach and a bush-based algorithm to solve decomposed subproblems. The condition of proportionality is imposed to obtain unique traffic flows for each class of vehicles on road links. In addition, a spatially constrained multivariate clustering technique is employed to construct origin/destination zones that are smaller than the coarse Freight Analysis Framework (FAF) zones, while maintaining reasonable computational burden for network traffic assignment. Model implementation in the U.S. shows that platooning could lead to 7.9% fuel saving among platoonable trucks in 2025 and a comparable increase in effective capacity of platoonable road links, which would account for 60% of rural interstate roads. The fuel saving and road capacity improvement translate into an annual cost reduction of $868 million for the U.S. intercity trucking sector and reduced road infrastructure investment needs worth $4.8 billion. Extensive sensitivity analysis further reveals that fuel saving of platoonable trucks increases with platoon size but decreases with inter-truck distance in a platoon. Fuel saving potential suggests that priority should be given to rural rather than urban roads in deploying platooning technologies. As expected, greater market penetration of platooning technologies means higher fuel saving and greater increase in effective road capacity.

Introduction

Emerging vehicle-to-vehicle wireless communication and control technologies enable truck platooning, i.e., a group of trucks travel in close longitudinal proximity of each other. The formation of a platoon brings two primary benefits. First, it reduces air drag experienced by platooning trucks, lowering their fuel consumption (McAuliffe et al., 2018). Field studies find that the reduction can be up to 10–15% (Bishop et al., 2017, Tsugawa et al., 2016). The consequent energy and economic implications are significant for the U.S. freight transportation sector where trucks carry about two thirds of the commodity flow (FAF4, 2019). In 2019, the annual fuel consumption by trucking in the country reached one billion barrels (EIA, 2019a). 10–15% of savings would mean large absolute amounts, which will become even greater in the decades to come as truck tonnage is projected to grow by 33% between 2019 and 2045 (FAF4, 2019).

In addition to fuel and emission reductions, truck platooning reduces road space use compared to trucks traveling individually, due to smaller inter-truck distance for trucks in a platoon. This increases the effective capacity of roads, which contributes to alleviating road congestion and improving transportation system performance (Tsugawa et al., 2016). Thus, truck platooning helps reduce the investment needs to add more road capacity which are estimated to be in the order of $10–30 billion annually for interstate highways alone in the U.S. (National Academies, 2019). Even very modest road capacity increase gained by truck platooning would mean huge savings in infrastructure investment.

It is thus not surprising that the promise of platooning has garnered increasing interest from industry, government, and the research community. In the growing body of academic literature (e.g., see Bhoopalam et al. (2018) for a recent review), two research problems are of relevance to the present study: the problem of platoon formation and the problem of platooning vehicle routing. For the former, Larson et al. (2015) explore ways to coordinate platoon formation using a network of controllers. The question of how individual trucks can cooperate to form platoons in a fuel-efficient manner is examined in Liang et al. (2016). Boysen et al. (2018) present a scheduling problem for truck platoon building along a single path. Zhang et al. (2017) formulate a platoon coordination and departure time scheduling problem under travel time uncertainty. Sun and Yin (2019) investigate the optimal platoon formation to maximize platooning benefit and then determine a mechanism to redistribute the benefits to incentivize vehicles to form and maintain the platoons. A model to optimize truck platoon formation at a platooning hub is proposed by Larsen et al. (2019). For the problem of platooning vehicle routing, Larsson et al. (2015) consider a graph routing problem to obtain the optimal platooning vehicle routing given a collection of starting points, destinations, and deadlines. Sokolov et al. (2017) explore simultaneously platoon coordination and vehicle routing.

Despite these existing efforts, some gaps remain in the literature. First, prior work has not looked into the network congestion effects of truck platooning, which also affect truck fuel use. Second, existing studies all assume that platooning vehicles schedule and announce their trip information (e.g., origin, destination, departure time, etc.) either before or during their actual trips. Different from scheduled platooning, spontaneous platooning may also be likely, especially in the U.S. given that the trucking industry is very fluid and fragmented (Medwell, 2016). Under spontaneous platooning, moving trucks match “on the go” with other trucks in close proximity to form platoons, without stopping or prior planning and scheduling (Bhoopalam et al., 2018). However, spontaneous truck platooning has not been well studied, in particular at the network level.

This paper attempts to make three contributions to the still scarce literature of system-level modeling of spontaneous truck platooning, with the U.S. national road network as the application context. For the first contribution, we propose a new modeling approach to characterize spontaneous truck platooning with coexistence of non-platooning vehicles at the network level, and a solution approach to solve the platooning-embedded multiclass network equilibrium. The model explicitly accounts for three interlocking relationships that arises from platooning. First, trucks may deviate from their original driving profile in order to be grouped with nearby trucks to form a platoon. This introduces platoon formation time (PFT). Second, truck fuel saving while platooning on a road link is modeled as a function of platoon characteristics (e.g., platoon size and inter-truck distance), PFT, and travel time on the link. Third, the effective road capacity increase due to truck platooning is modeled as a function of inter-truck distance, platoon size, and the proportion of platooning trucks in total traffic. PFT and effective road capacity increase on a road link are derived as functions of the platooning truck flow on the link, which in turn depends on PFT and effective road capacity. To equilibrate the relationships, an algorithm is proposed which involves a diagonalization approach and Dial’s bush-based algorithm to solve decomposed subproblems. In addition, the condition of proportionality is imposed to obtain unique traffic flows for each class of vehicles on road links, which is necessary to calculate fuel savings due to truck platooning.

The second contribution of the paper lies in the employment of a spatially constrained multivariate clustering technique to construct zones, which is necessary for truck trip generation/attraction. Given that the model is to be applied to the U.S., existing zoning systems are either too coarse to be compatible with the national road network, thereby causing significant inaccuracy in traffic assignment, or too detailed that traffic assignment would become computationally expensive. The proposed clustering technique aggregates those small zones in the more detailed zoning system into larger (but not too large) zones. In performing the clustering, similar small zones are placed in the same larger zone whereas dissimilar small zones are placed in different larger zones. Dissimilarity/similarity are measured using multiple socioeconomic attributes and distance information. In doing so, every aggregated zone is further ensured to be contiguous which is desired for truck trip generation/attraction. The clustering proceeds by first constructing a minimum spanning tree (MST) which includes all small zones to be clustered, and then partitioning the MST guided by a neighborhood search-based heuristic.

As the third contribution of the paper, a number of insights are generated from implementing the model in the U.S. The investigation of truck platooning in an overall national transportation system, to our knowledge, has not been done before in the literature. Through extensive numerical analysis, we estimate that in the baseline scenario for year 2025, platoonable trucks could save 7.9% fuel with the effective capacity of platoonable road links increased by a comparable percentage. This translates to close to one billion-dollar cost reduction in the U.S. truck industry annually and reduced road infrastructure investment needs worth nearly five billion dollars. We further find that fuel saving potential is quite sensitive to platoon size, inter-truck distance in a platoon, and market penetration of platooning technologies. Rural roads have greater potential than urban roads in reaping platooning benefits. These findings, while grounded on assumptions, are helpful in informing future decision- and policy-making for truck platooning technology investment and deployment.

The remainder of the paper is organized as follows. The next section presents the platooning-embedded multiclass network equilibrium model and the solution algorithm to determine network flows of platooning trucks in coexistence of non-platooning vehicles. The spatially constrained multivariate clustering technique to generate zones for truck trip generation/attraction is developed in Section 3. Section 4 describes in detail data preparation. Results are presented and analyzed in Section 5. Conclusions and directions for future research are provided in Section 6.