Internet of Things (IoT) devices have been rapidly deployed and data generation grows fast. Therefore, edge computing has become a crucial mechanism to bring resources close to the users’ locations However, task scheduling optimization for edge computing environments faces significant challenges, particularly in the scenario of energy consumption and resource utilization. Conventional task scheduling algorithms usually fail to efficiently address above challenges. In this paper we propose a novel approach to energy-efficient task optimization for big data analysis workflows in edge computing based on the Firefly Algorithm (FA). The FA dynamically allocates tasks across heterogeneous edge nodes to minimize energy consumption while maintaining high performance. Experiments show that the FA-based method reduces energy consumption compared to conventional meta-heuristic algorithms, Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO). The results highlight the FA's potential in enhancing sustainability and operational efficiency in edge computing.