PREDICT - Predictive Consolidated Transportation
The PREDICT project aims to increase resource efficiency and transport efficiency via predictive management of mixed-sized vehicles fleets for consolidated/shared transportation of people and goods.
Purpose
Increase resource efficiency and transport efficiency via predictive management of mixed-sized vehicles fleets for consolidated/shared transportation of people and goods.
The project activities related to the projects aims are:
(1) Deep neural network based travel time and consolidated transport demand predictions and road network based propagation for routing; (2) optimized predictive fleet management (i.e., dispatch and re-positioning strategies/control) for mixed sized fleets using deep reinforcement learning, and (3) fleet management simulations based on 100s of millions of taxi request data in NYC.
The project is funded by ITRL. For more information, contact Gyözö Gidofalvi or Jesper Provoost .