Abstract: Optimising pilot scheduling can reduce working hours and labour costs for pilots, while offering new approaches to pilot training. This paper addresses the scheduling of pilots for inbound and outbound vessels in a one-way waterway by establishing a joint optimisation model, with the aim of minimising total vessel waiting time, total pilot working hours and labour costs. It proposes a two-stage heuristic algorithm that combines an imperialist competitive algorithm, an adaptive evolutionary strategy, and a variable neighbourhood search (VNS) algorithm. Case studies and comparative experiments demonstrate the effectiveness of the model and algorithm. Compared with the traditional first-come, first-served (FCFS) rule, the proposed approach reduces pilots' average working hours and labour costs by 17.4% and 12.4%, respectively. Therefore, integrating the model and algorithm into pilot training programmes or operational guidelines could enhance coordination among port departments, improve organisational efficiency and service quality, and support the development of green and smart ports.