Workload management is an important component of Crew Resource Management, which is now a mandatory training and screening requirement for all airline pilots worldwide. A pilot’s workload per flight typically varies from time to time, influenced by the flight phase (climb, cruise, descent), weather conditions, and other factors such as mental state and skill level. A competent pilot will strive to manage the workload by conducting tasks such as checklists and communicating with ATC timely during each flight phase to avoid creating situations where the workload is extremely high. However, it is not easy for pilots to understand and predict changes in their own workload. In this study, as a new method to support the pilot’s workload management during flight, we construct a model to estimate the pilot’s workload at t + Δt from various data at time t in flight. Furthermore, we propose and validate a system in which the workload after Δt output by this model is always displayed in the cockpit at t.