Dan McCabe, Ph.D. Transportation, Energy, Operations Research

My Research

Ongoing Work at National Laboratory of the Rockies (NLR)

At NLR, I’m based in the Computational Science Center and collaborate with researchers throughout the lab and beyond on emerging problems in transportation and computing. Some active research projects are described below.

INtermodal Freight Optimization For a Resilient Mobility Energy System (INFORMES)

  • We’re building INFORMES, a national-scale model for analyzing multimodal freight transportation. I’m using SimPy to develop a data-driven discrete-event simulation model at the core of this project. I work closely with our team at NREL and collaborators from several other institutions to model container freight movements throughout the country, simulating shipments over road, rails, and water and transfers at intermodal facilities. The INFORMES tool will enable stakeholder across both public and private sector organizations to evaluate freight system reliability, costs, and energy efficiency across a wide range of future scenarios.

Transit Vehicle Selection Optimization and Total Cost of Ownership Evaluation

We’re working with the Federal Transit Administration (FTA) to develop a public app to help transit agencies decide which buses to include in their fleets based on data-driven projections of energy usage and total cost of ownership for a wide range of bus types—including 40- and 60-foot models of diesel, hybrid, fuel cell, battery-electric, and CNG powertrains.

For this project, I’ve developed and streamlined a pipeline to gather GTFS data on all American transit agencies and predict the energy consumption of different bus models on each trip defined in this data, taking into account expected vehicle speeds, elevation differences, and weather conditions. This process relies on NLR’s existing RouteE-Powertrain tool, the same technology that powers Google Maps’ energy-aware routing capability. We then process these results into inputs for NLR’s Transportation Technology Total Cost of Ownership (T3CO) model to estimate lifetime costs of each vehicle model given each agency’s unique service patterns.

The core analysis capability of this project is available in the open-source routee-transit GitHub repository.

Transit System Resilience with Mixed Fleets

  • I’m developing a simulation platform that performs detailed simulation of transit bus fleets with mixed powertrain types. The simulation models typical operations as rell as key resilience scenarios, such as charging infrastructure outages and vehicle failures, as well as mitigation measures to enhance transit system resilience. The software will help transit agencies design and operate their bus systems for maximum performance, efficiency, and reliability.

NLR Modeling, Simulation, and Optimization Capability (MSOC)

  • Designed and presented a tutorial session for NLR researchers on using GitHub Copilot AI assistant with Visual Studio Code. The session was attended by over 150 NLR research staff.
  • Completed an edX course on quantum computing to support future NLR research in this area of increasing priority
  • Supported usage of high-performance computing across the lab, especially Kestrel supercomputer

PhD Research

Buses charging in Gothenburg, Sweden. In front of the fast electric bus chargers at Heden terminal, Gothenburg, Sweden.

My PhD research focused on battery-electric buses (BEBs). Transit agencies across the U.S. (as well as the rest of the world) have ambitious plans to reduce their emissions and BEBs will be an important technology to support that goal. However, they also present challenges to bus operators due to limited driving ranges and slow recharging times in comparison with buses that use liquid fuels. My work on BEBs consists of two main thrusts:

1) Optimization models for both long-range planning and daily operations of transit systems using BEBs. This includes planning for recharging infrastructure and scheduling charging during the day. These problems are simple to understand and state, but solving them turns out to be very computationally challenging.

2) Software tools to facilitate running these models and visualizing their results. This includes one public-facing tool, the Zero-Emission Bus Range & Recharging Assessment (ZEBRA). This public web app is compatible with most American transit agencies’ public GTFS data and processes data on vehicle blocks (daily trip assignments for each bus) to determine range requirements and potential recharging needs.

Technical Expertise

The technical side of my work centers on optimization, particularly mixed-integer linear programming. My experience with operations research dates back to my final year of undergrad in 2016-17 (thank you to my awesome professor Susan Martonosi for getting me into this subject!) and I’ve been working in this field ever since, both at UW and in my previous job at Pacific Northwest National Laboratory. My expertise includes:

  • Developing optimization models across a variety of domains, but especially transportation, power systems, and network modeling.
  • Integer programming algorithms, including:
    • Branch-and-cut (and its application in modern solvers using callbacks)
    • Decomposition methods such as Benders Decomposition and Lagrangian Relaxation
  • Solving optimization models with modern software tools, especially. I’m particularly familiar with the solver-agnostic pyomo library and Gurobi’s gurobipy, both of which I have used extensively. I have also used MATLAB, GAMS, and AMPL for optimization work.
  • Data processing and visualization tools to help with optmization inputs and outputs. I have several years of experience with major Python libraries like pandas, numpy, matplotlib, and plotly.

Publications

McCabe, D. Computational Tools for Battery-Electric Bus Systems: From Infrastructure Planning to Daily Operations. Ph.D. dissertation, University of Washington, 2024. Available here via Proquest.

McCabe, D., X.J. Ban., and B. Kulcsár. Recharging Scheduling for Electric Buses with Exact Delay Propagation. Under revision for Transportation Research Part E: Logistics and Transportation Review. arXiv preprint available at https://arxiv.org/abs/2403.17527.

McCabe, D. and X.J. Ban. Optimal Locations and Sizes of Layover Charging Stations for Electric Buses. Transportation Research Part C: Emerging Technologies 152 (2023): 104157.

McCabe, D. Selecting Layover Charging Locations for Battery-Electric Buses: Mixed-Integer Linear Programming Models. Master’s thesis, University of Washington, 2021. Available at: http://hdl.handle.net/1773/47413

See my CV for a full list of publications and presentations.

Two BEBs plugged in to charge.