arXiv subjects

Anne Driscoll

Publications and source records attributed to Anne Driscoll.

3 recordsLinked to original sources

arXiv2026

Using large scale GPS data to reveal EV driver activity patterns beyond charging sessions

Accurate insights into electric vehicle (EV) driver behavior are essential for long-term infrastructure planning, grid management, and understanding downstream economic impacts, yet individual level data on EV mobility remains limited. Here, we develop a scalable framework to infer EV ownership and charging behavior from passively collected, high-resolution mobility traces covering over 760,000 drivers across four major U.S. metropolitan areas. We identify likely EV drivers based on distinctive visitation patterns to charging stations and gas stations, frequency of visits, and daily travel behavior, and calibrate cohort size using aggregate EV registration statistics. The resulting EV cohort closely matches official registration data at the zip code level and exhibits charging patterns consistent with independent, charger level benchmark datasets, providing external validation of the inferred population. Leveraging this inferred cohort, we reconstruct charging events and associated activity patterns to examine how EV drivers interact with surrounding urban amenities. Compared to non-EV drivers, EV drivers exhibit systematically higher visitation rates to nearby cafes and restaurants during charging sessions, revealing significant economic spillover effects. Furthermore, we find EV drivers exhibit trip bundling behavior, visiting more POIs over less time and distance on days where they charge versus all other days. These patterns are not observable in conventional charging session data, which lack behavioral context beyond the charging event itself. Our results demonstrate the potential of using mobility data to enable a richer, behaviorally grounded understanding of the off-plug needs of EV drivers, providing a foundation for optimizing charging infrastructure deployment and co-locating complementary urban amenities in an increasingly electrified transportation landscape.

physics.soc-ph↗

arXiv2021

SustainBench: Benchmarks for Monitoring the Sustainable Development Goals with Machine Learning

Progress toward the United Nations Sustainable Development Goals (SDGs) has been hindered by a lack of data on key environmental and socioeconomic indicators, which historically have come from ground surveys with sparse temporal and spatial coverage. Recent advances in machine learning have made it possible to utilize abundant, frequently-updated, and globally available data, such as from satellites or social media, to provide insights into progress toward SDGs. Despite promising early results, approaches to using such data for SDG measurement thus far have largely evaluated on different datasets or used inconsistent evaluation metrics, making it hard to understand whether performance is improving and where additional research would be most fruitful. Furthermore, processing satellite and ground survey data requires domain knowledge that many in the machine learning community lack. In this paper, we introduce SustainBench, a collection of 15 benchmark tasks across 7 SDGs, including tasks related to economic development, agriculture, health, education, water and sanitation, climate action, and life on land. Datasets for 11 of the 15 tasks are released publicly for the first time. Our goals for SustainBench are to (1) lower the barriers to entry for the machine learning community to contribute to measuring and achieving the SDGs; (2) provide standard benchmarks for evaluating machine learning models on tasks across a variety of SDGs; and (3) encourage the development of novel machine learning methods where improved model performance facilitates progress towards the SDGs.

cs.LG↗

arXiv2020

Using satellite imagery to understand and promote sustainable development

Accurate and comprehensive measurements of a range of sustainable development outcomes are fundamental inputs into both research and policy. We synthesize the growing literature that uses satellite imagery to understand these outcomes, with a focus on approaches that combine imagery with machine learning. We quantify the paucity of ground data on key human-related outcomes and the growing abundance and resolution (spatial, temporal, and spectral) of satellite imagery. We then review recent machine learning approaches to model-building in the context of scarce and noisy training data, highlighting how this noise often leads to incorrect assessment of models' predictive performance. We quantify recent model performance across multiple sustainable development domains, discuss research and policy applications, explore constraints to future progress, and highlight key research directions for the field.

cs.CY↗