Aye-Aye Optimizer (AAO): A Bio-Inspired Metaheuristic Algorithm Based on the Percussive Foraging Strategy

Abstract Complex optimization problems frequently arise in engineering design, control, machine learning, and scientific computing, where the objective landscapes are often nonlinear, multimodal, constrained, and high-dimensional. Efficiently solving such problems is important because the quality of the obtained solutions directly affects design reliability, computational cost, and decision-making performance. Although numerous metaheuristic algorithms have been developed to address these challenges, maintaining a proper balance between global exploration and local exploitation remains a key difficulty, particularly in complex search spaces with multiple local optima. To address this issue, this paper introduces the Aye-Aye Optimizer (AAO), a bio-inspired metaheuristic algorithm that mimics the percussive foraging strategy of the Aye-Aye (Daubentonia madagascariensis), a nocturnal lemur native to Madagascar. In nature, the Aye-Aye uses rhythmic tapping to detect hidden prey beneath tree bark, followed by localization using auditory cues and final extraction with its elongated middle finger and strong incisors. Inspired by this behavior, the AAO models the optimization process through three sequential mechanisms: exploratory percussion for broad global exploration, focused localization for refining promising regions, and local exploitation for intensifying the search around high-quality candidate solutions. The transition among these mechanisms is controlled by three parameters, namely alpha, beta, and gamma, which regulate the balance between exploration and exploitation during the optimization process. The performance of the proposed AAO is evaluated using 23 classical benchmark functions, CEC2022 test functions, and several constrained engineering design problems. The results indicate that AAO provides competitive performance in terms of solution accuracy, convergence behavior, and robustness compared with selected well-established metaheuristic algorithms. These findings suggest that the percussive foraging-inspired search mechanism can serve as a promising framework for developing population-based optimization algorithms.