Dynamic Bayesian network (DBN) theory provides a valid tool to estimate the risk of disruptions, propagating along the supply chain (SC), i.e. the ripple effect. However, in cases of data scarcity, obtaining perfect information on probability distributions required by the DBN is impractical. To overcome this difficulty, a new robust DBN approach is, for the first time, proposed in this study to analyse the worst-case oriented disruption propagation in the SC. This work considers an SC with multiple suppliers and one manufacturer over several time periods, in which only probability intervals of the suppliers' states and those of the related disruption propagations are known. The objective is to acquire the robust performance of risk estimation, measured by the worst-case probability in the disrupted state for the manufacturer. We first establish a nonlinear programming formulation to mathematically materialise the proposed robust DBN, which can be used to solve small-size problems. To overcome the computational difficulty in solving large-size problems, an efficient simulated annealing algorithm is further designed. Numerical experiments are conducted to validate its efficiency.

Assembly line worker assignment and balancing problem (ALWABP) is an important research topic originated from sheltered work centres for disabled, in which workforce is assumed to be heterogeneous due to their disabilities. Since the employment of disabled workers may sustain higher absenteeism rates due to their health, especially under COVID-19, employing temporary workers to fill labour shortage is a crucial issue. In addition, in practice, the movement of workers between stations on assemble lines can increase the flexibility of worker assignment. In this study, we investigate a new risk-averse ALWABP with uncertain disabled worker availability, limited temporary workers and moving workers. The objective is to minimise the risk-averse weighted sum of the cycle time and the number of employed temporary workers. For the problem, a risk-averse two-stage stochastic programming model is formulated. The first stage assigns specific tasks (called fixed tasks) to stations, while the second stage assigns workers and remaining tasks (called flexible tasks) to stations. A genetic algorithm combining K-means clustering approach and variable neighbourhood search (GAKV) is designed. Experiment results show the superiority of the GAKV in terms of solution quality and computation time compared with sample average approximation (SAA). In addition, managerial insights are drawn.

Cargo (laden container) routing and empty container repositioning are crucial components in liner shipping, which are closely relevant. However, most existing works focus on cargo routing and empty container repositioning separately, which cannot optimize the global shipping cost. In addition, (1) appropriate repacking operations for laden containers can curtail the total handling costs, and (2) the perfect information for uncertain laden and empty container demands may be difficult to obtain due to insufficient historical data. This work investigates an integrated cargo routing and empty container repositioning problem with repacking operations in which only partial information about uncertain laden and empty container demands is known. The objective is to minimize the total costs including repacking operation cost and laden and empty container transportation costs. For the problem, a new chance-constrained programming model based on moment-based ambiguous sets is formulated. Then, four distribution-free solution approaches are adopted to solve the investigated problem. Numerical experiments are conducted to compare the proposed methods.