WB6 - Mining and Natural Resources 2

We consider a wood-supply planning problem for a pulp and paper mill operating within a complex supply chain. Decision makers must forecast requirements in advance to ensure sufficient volume and avoid costly disruptions. We formulate a Mixed Linear Programming model that jointly optimizes transportation and inventory decisions, and present preliminary computational results.

This study develops an optimization model for mixed-fleet and charging infrastructure planning in forest inter-mill transportation networks comprising electric, hybrid, and diesel trucks. The model selects truck-route assignments, Truck type, and identifies the recharging locations. The model optimizes the total transportation cost while ensuring that all transports required are done.

Keywords:

Fleet optimization; Mixed fleet; Forest transportation; Electric vehicles; Hybrid vehicles; Diesel vehicles

Abstract:
Autonomous fleet coordination in bidirectional single-lane corridors remains challenging due to collision and deadlock risks. This study proposes a digital twin–based decentralized framework enabling proactive conflict detection and rerouting. Simulation results show global visibility reduces mean cycle delay by 20% and detour distance by 63% compared to driver-like arbitration policies.
Keywords: Conflict Detection, Underground Mining, Fleet Management System.

Fossil fuel-based transportation technologies are increasingly being electrified or hybridized to reduce environmental impact and leverage cheaper energy sources. The automotive sector has demonstrated that such transitions can reduce greenhouse gas emissions while maintaining cost-effectiveness and system performance. In contrast, small- to medium-scale fishing vessels have been slower to adopt electrification due to strict operational constraints, diverse usage profiles, narrow profit margins, and high upfront investment costs. Despite these challenges, substantial environmental and economic potential exists if the design and operations (D&O) of these vessels are jointly optimized. This work proposes a bi-level optimization model that simultaneously addresses D&O using experimental data collected from fishing vessels in Québec, Canada. The upper level seeks to maximize the return on investment over multiple scenarios through a deterministic fractional linear programming formulation. It incorporates actual naval equipment, e.g., marine electrical propulsion systems and battery energy storage solutions, and economic assumptions provided by naval architects, ensuring practical relevance. The lower level ensures optimal energy management system (EMS) of the vessel given a design. The problem is solved as a single-level problem via an exact reformulation based on the Karush-Kuhn-Tucker conditions and disjunctive constraints with off-the-shelf mixed-integer linear solvers. As such, the proposed framework assesses the design under optimal operations that depends on the deployed EMS strategy, e.g., maximizing the internal combustion engine’s (ICE) efficiency, keeping the RPM in an optimal range or minimizing the ICE usage. The configuration of the problem makes it so multiple EMSs can easily be integrated and tested with out-of-sample data after optimal D&O with real life step size.

Sugarcane production, a key global industry, requires tight coupling between harvesting and milling. Disruptions in harvesting, transportation, and milling may destabilize these systems. In response, this study proposes a MILP model to rebalance the flow of cane from harvest fronts to mills following such disruptions, aiming to reduce overall costs.