Machine learning & AI

A new machine learning approach to forecasting future transportation demand

New research from University College Cork (UCC) and Columbia University will increase the precision of forecasting future demand for passenger and freight transportation, which together account for 20% of the world’s greenhouse gas emissions.

According to UN estimates, by the year 2050, there may be about 9 point 7 billion people on the planet, up from 7 point 7 billion in 2019. There will probably be more demand for transportation services as a result of the growing population and economy.

For climate policy, lowering emissions related to transportation remains a significant challenge. Up until now, tasks involving forecasting transport demand were completed by simulating demand or using regression-based analysis. Worldwide nations can now more precisely predict future transportation needs thanks to UCC and Columbia University research.

“This research provides insights into the development of a novel machine learning architecture that improves the accuracy of transport energy service demand estimation. The unique machine learning architecture and its benefits for the energy modeling community are quantifiable and applicable to other fields.”

Siddarth Joshi, who led this research as part of his Ph.D. in Energy Engineering at UCC .

This study, which was published in Scientific Reports, introduces TrebuNet, a brand-new, cutting-edge machine learning technique. The outcomes show that, in comparison to both conventional regression techniques and more recent cutting-edge neural network and machine learning techniques, this new TrebuNet architecture achieves superior performance. The enhancements include regional demand projections for all modes of transportation at short-, long-, and intermediate-term time scales.

Siddarth Joshi, who directed this study as a component of his Ph.D. “This study provides insights into the development of a novel machine learning architecture that increases the accuracy in the estimation of transport energy service demands,” said Energy Engineering at UCC. The advantages of the cutting-edge machine learning architecture are quantifiable for the community of energy modeling and are transferable to other disciplines.”.

According to Brian Gallachóir, a professor of energy engineering at UCC, “accurate transport demand projections are important not only for energy system models and climate policy, but they also serve as the foundation for understanding the future direction of global energy markets.”

Dr. According to James Glynn, Senior Research Fellow at Columbia University, “This new method demonstrates innovation in data analytics and energy systems modeling to address a weakness in understanding the outlook within energy system models for new applications of deep learning. This aids in removing uncertainty from decarbonization pathways.”.

“Urgent climate action is required to decarbonize transportation in accordance with global net-zero 2050 targets. In order to give decision-makers the tools and evidence-based research they need to design climate policy, collaboration between Columbia SIPA and UCC has led to new approaches in data science and energy systems modeling.

More information: Siddharth Joshi et al, A deep learning architecture for energy service demand estimation in transport sector for Shared Socioeconomic Pathways, Scientific Reports (2023). DOI: 10.1038/s41598-023-30555-6

Topic : Article