HomeMachine LearningMachine Learning NewsEstimating future demand in the transport sector using ML

Estimating future demand in the transport sector using ML

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

According to UN predictions, there will be about 9.7 billion people on the planet in 2050, up from 7.7 billion in 2019. The population increase and economic expansion will probably result in a rise in demand for transportation services.

For climate policy, lowering emissions connected to transportation remains a significant challenge. Up until recently, jobs involving the projection of transport demand were completed using demand simulation or regression-based analysis. Now that this UCC and Columbia research has been completed, nations all across the world will be able to predict future transportation needs with more accuracy.

This study, which was published in Scientific Reports, introduces TrebuNet, a brand-new, cutting-edge machine learning technique. The outcomes show that this novel TrebuNet design outperforms both established regression techniques and more modern, cutting-edge neural network and machine learning techniques. The enhancements include regional demand projections for all forms of transportation at short-, long-, and intermediate-term time scales.

As part of the doctoral program in Energy Engineering at UCC, Siddarth Joshi, the project’s principal investigator, said, This study provides insights into construction of a novel machine learning architecture that boosts the accuracy in the assessment of transport energy service demands. The advantages of this cutting-edge machine learning architecture can be used across disciplines and are quantifiable for the energy modelling community.

According to Brian Gallachóir, professor of energy engineering at UCC, accurate transportation demand projections are important for energy system models and climate policy, as well as serving as the foundation for understanding the future trajectory of global energy markets.

According to Dr. James Glynn, Senior Research Fellow at Columbia University, This novel method displays innovation in data analytics and energy systems modelling to address a deficiency in comprehending the outlook inside energy system models for new applications of deep learning. This helps us eliminate doubt about decarbonization processes.

Urgent climate action is required to decarbonize transportation in accordance with global net-zero 2050 targets. New methodologies in modelling energy systems and data science are being developed as a result of collaboration between Columbia SIPA and UCC, giving decision-makers the resources and evidence-based research they need to establish effective climate policies.

Source link

Most Popular