Iterative Function Allocation of Single Pilot Operations Based on Biclustering Method

Abstract

The successful transition from a two-pilot to a single-pilot mode cannot solely rely on simply augmenting advanced automation in the cockpit. The degree of support provided by functions at different levels of automation varies, and introducing new functions simultaneously brings new demands for the single pilot, leading to imbalanced workload. To address these shortcomings and enhance the rationality of function allocation under workload constraints, this paper introduces an iterative function allocation method. Initially, the functions are categorized and classified into different levels of automation to obtain the initial function allocation scheme. Subsequently, a novel bicluster mining algorithm called K-Miner is proposed to identify elements requiring iterative design. As a case study, the multitask surface scenario in SPO is considered. By applying the proposed iterative design method, the task performance of the single pilot is improved, the workload of the single pilot is effectively reduced, and situational awareness is improved to a certain extent.