• Energy Management System
• Green Logistics/Warehousing
• Mixed Integer Linear Programming
• Sequencing and Routing Problem

Amidst the growing awareness of environmental sustainability within the logistics sector, warehouse activities face increasing scrutiny due to their contribution to greenhouse gases and carbon emissions. Driven by the decarbonisation trend, renewable energy sources like rooftop photovoltaic panels, together with a battery storage system, are being employed to move towards zero-carbon electricity for warehouse operations. This thesis investigates the integration of green practices within a warehouse context, focusing on extending the Green Sequencing and Routing Problem, which entails determining the optimal sequence of locations to visit within a warehouse, for storage and retrieval operations, utilising a mixed fleet composed of both electric vehicles (equipped with lithium-ion batteries) and conventional vehicles (with internal combustion engines), with rooftop photovoltaic panels and a battery storage, in addition to the conventional power grid. Specifically, this study explores the proposal of an Energy Management System aiming to optimise energy management among the rooftop photovoltaic panels, battery storage system and conventional power grid, in order to satisfy the energy requirements of the warehouse and, in particular, of the electric vehicles in charge of storage and retrieval operations within the warehouse. To address the proposed problem, a Mixed Integer Linear Programming model is presented, which extends a model from the literature addressing the Green Sequencing and Routing Problem without energy management considerations. Preliminary computational findings are also presented to demonstrate the feasibility of the proposed approach. This research has embraced a collaboration with some engineers from University of Genoa, particularly regarding the definition of the energy management system and the setting of data related to the energy parameters of the mathematical model, thereby facilitating the testing of the model in a realistic computational environment, although of reduced size with respect to real instances.