In order to meet item demands, end-of-life (EOL) product and subassembly ordering and disassembly schedules are determined by disassembly lot sizing, which is the subject of this study. We take into consideration a stochastic version with undetermined ordering lead time (OLT). In this case, OLT stands for the amount of time that passes between placing and receiving an order (we can only order EOL products). Throughout the planning horizon, scenarios are used to model the stochasticity. The objective is to reduce the expected total of setup, purchasing, inventory, and backlog expenses. This is achieved by expressing the problem as a two-stage mixed integer linear programming (2S-MILP) model across all potential scenarios. The 2S-MILP is unsolvable since it is predicated on every scenario conceivable. A Simultaneous Backward Reduction approach is proposed to make it tractable. To confirm the suggested method’s efficacy, it is assessed in a variety of environments.

This paper studies the assembly line balancing problem with collaborative robots in light of recent efforts to implement collaborative robots in industrial production systems under random processing time. A stochastic version with uncertain human processing time is considered for the first time. The issue is defined by the potential for simultaneous human and robot task execution at the same workpiece, either in parallel or in collaboration. We provide stochastic mixed-integer programming based on Monte Carlo sampling approach for the balancing and scheduling of collaborative robot assembly lines for this novel issue type. In order to minimise the line cost including fixed workstation operating costs and resource costs caused by exceeding cycle time, the model determines both the placement of collaborative robots at stations and the distribution of work among humans and robots.

Parallel machine scheduling problems are among the most studied scheduling problems in the literature. We study the unrelated parallel machine scheduling problem with release dates and machine-and sequence-dependent setup times to minimize the makespan. We introduce three heuristics, five local search methods and three metaheuristics for the problem: the Late Acceptance Hill Climbing and two variants of Simulated Annealing. Furthermore, we introduce a three-set 1620-instance benchmark in which the number of jobs and machines, release dates, processing and setup times are generated according to existing procedures. The proposed approaches are analyzed and compared on the proposed benchmark. We compare the proposed heuristics and derive the best heuristic to be used to generate the initial solution of the metaheuristics. Moreover, we show that the different metaheuristics are efficient, each performing best on one of the sets.

An examination timetabling problem taken from real world universities was proposed at the International Timetabling Competition (ITC2007). The aim was to establish a common base for comparing different solution approaches. This paper presents new preprocessing methods that disclose hidden constraints and significantly increase the number of new edges that can be added to the conflict graph. Results show that the size of the maximum clique of the obtained conflict graph has been more than doubled for two instances as a result of our preprocessing. These larger cliques mean that instances can be analyzed in advance of a solution and end users gain useful information for making decisions. In addition, we have looked at the different criteria that compose the objective function, in order to provide more useful insights into the difficulty of problems in practice. We propose new integer programming formulations using clique inequalities to compute optimal solutions for 4 criteria and to obtain lower bounds for the 3 others. Results are presented and discussed for all the benchmark instances.