Shahrooz Shahparvari

Bushfire evacuation transport plans are committed on uncertain demand. Before warnings, pickups and shelter places are fixed, an incident team can task only a few sensing channels (aerial scans, mobility data, road counters, door-knock crews), each with a cost, latency and noise. We ask which sensing changes the decision. We formulate the problem as a partially observed Markov decision process whose terminal decision is a two-stage capacitated zone-to-pickup and shelter assignment, and rank observations by the decision regret they remove per unit of sensing cost, counting a signal only if it lands before the plan is due. On a separable subclass, variance-minimising sensing can capture no decision value, and with unit costs latency turns acquisition into deadline scheduling, where a greedy rule discarding late signals keeps at least half the attainable value. In a coupled experiment, ignoring landing times costs 9 to 23% in regret. In a 2019–20 East Gippsland hindcast with assumed shelter capacities and simulated sensing, three observations remove 99% of modelled provisioning regret out of sample, against 88% for variance sensing, and 66% against 58% when onesignal in four fails; with slack shelters no observation has value. Across 1,965 Victorian localities and 23 fire seasons, from pre-season data only, about 111 localities a season are in the balance; observing them at budgets of 10 and 20 lowers realised regret, mainly by avoiding warnings where fire did not arrive. The framework values a sensing stream by the transport plans it can change, not the uncertainty it removes.

Wildfire evacuation is a routing problem, but the road network it runs on does not hold still. Links close as the fire advances, and they reopen only when suppression crews, scheduled by a separate agency, clear them. Evacuation flow therefore depends on the timing and order of suppression. We study evacuation on such a network, where link availability is endogenous and contested: suppression access and evacuation egress claim the same corridors at once. This generalises contraflow and dynamic-capacity evacuation models that treat time-dependent capacity as given. The model is a multi-stage stochastic mixed-integer linear program. It coordinates firefighting agencies, police, and emergency management authorities across resource allocation, operational scheduling, road clearance, evacuationrouting, shelter assignment, contraflow, and blocked-link reopening, under wildfire propagation, uncertain operating conditions, and time-dependent road accessibility. To solve it, a Parallel Branch-and-Fix Coordination procedure returns a certified lower bound and feasible solutions through scenario clustering and Lagrangian relaxation of nonanticipativity.We test it on the 2023 McDougall Creek wildfire in British Columbia, Canada, instantiated on a real road network. The value of coordination proves conditional. When a single resource such as shelter capacity binds, coordinating the two operations gains almost nothing. When suppression order gates the reopening of evacuation corridors, it cuts the un-evacuated population sharply. Sensitivity analyses over travel time, time windows, resourceavailability, and the shortfall penalty identify which parameters drive the plan. The lesson is a bottleneck-shift condition: coordinating suppression and evacuation pays only when suppression timing, not a fixed capacity, limits how many people reach safety.

The traditional patrol systems are unable to cover all regions of the forest and pastures due to the difficulty of crossing some routes. This study aims to create a new bi-objective mathematical model for forest fire protection by proposing a new coordinated patrolling system. Drones serve as a useful tool for monitoring impassible areas, while trucks serve as mobile depots for them by considering uncertainty in the time and cost parameters when implementing a scenario generation method. An Accelerated Benders' decomposition (ABD) is developed to find the optimal solution for each scenario. To evaluate the model robustness, risk-averse and min-max relative regret robust optimization approaches are analyzed. Finally, the column and constraint generation (CCG) method, a highly efficient method for solving scenario-based problems, has been developed. The results show that the robust model outperformed the deterministic model in all scenarios. The CCG approach solves the specified risk-averse and min-max relative regret robust optimization models in a reasonable time for large sizes and scenarios. The proposed solution framework is hence capable of optimally planning for any proposed used truck-and-drone coordinated problem.

Purpose: This paper designs an optimal closed-loop supply chain network with an integrated forward and reverse logistics to examine the possibility of remanufacturing end-of-life (EoL) ships.Design/methodology/approach: Explanatory variables are used to estimate the number of EoL ships available in a closed-loop supply chain network. The estimated number of EoL ships is used as an input in the model and then it is solved by a mixed-integer linear programming (MILP) model of the closed-loop supply chain network to minimise the total logistic costs. A discounted payback period formula is developed to calculate the length of time to recoup an investment based on the investment's discounted cash flows. Existing ship wrecking industry clusters in the Western region of India are used as the case study to apply the proposed model.Findings: The MILP model has optimised the total logistics costs of the closed-loop supply network and ascertained the optimal number and location of remanufacturing for building EoL ships. The capital and variable costs required for establishing and operating remanufacturing centres are computed. To remanufacture 30 ships a year, the discounted payback period of this project is estimated to be less than two years.Practical implications: Ship manufacturing businesses are yet to re-manufacture EoL ships, given high upfront capital expenditure and operational challenges. This study provides management insights into the costs and benefits of EoL ship remanufacturing; thus, informing the decision-makers to make strategic operational decisions.Originality/value: The design of an optimal close loop supply chain network coupled with a Bayesian network approach and discounted payback period formula for the collection and remanufacturing of EoL ships provides a new integrated perspective to ship manufacturing.

Sanitary waste disposal and site selection for establishing landfills are challenging problems for environmental planners. This paper aims to take environmental, socio-economic, geological, geomorphological, hydrological and ecological factors into consideration to provide a decision support framework for landfill siting. Analytical hierarchy process (AHP) and Decision Making Trial and Evaluation Laboratory (DEMATEL) are coupled to develop an efficient multi-criteria decision-making method to be utilized in a Geographic Information System (GIS) environment for evaluating the suitability for landfill siting. As the first attempt to employ DEMATEL effectively in a landfill site selection problem, the proposed method is tested with landfill siting scenarios in New South Wales (NSW), Australia. Regional analysis is also performed to identify the potentially most suitable statistical divisions for landfill siting in NSW. The top two ranked zones covering 0.7% and 22% of the study area, respectively, are considered as the optimal areas for establishing landfills, while the bottom two ranked zones are not recommended for further consideration. Further detailed analysis is also conducted on the existing landfills, which shows that 1.0% and 37.0% of them are ranks 1 and 2, respectively. The scenario-based analysis implies that, among the contributing factors; geological and economic factors are highly important.