Under “double carbon” policies, the shift towards bus electrification is unavoidable but is hindered by operational challenges due to electric buses’ limited range. These challenges are intensified by heterogenous bus fleets, inadequate charging facilities, and depot limitations. This paper introduces a pragmatic approach to address the multi-depot electric bus scheduling problem, comprehensively addressing all aforementioned complexities. A tailored adaptive large neighborhood search (ALNS) is designed to solve the problem, wherein several ingenious accelerating techniques are integrated to generate a high-quality solution within an acceptance time frame, including: 1) relaxation of hard constraints to expedite iterations; 2) specialized operators designed to enhance the diversity of neighborhood spaces; 3) the meticulous setting of operator scores to prioritize operators with higher efficiency. A small-scale case study is conducted to compare the tailored ALNS with the commercial solver. The results substantiate the superior performance of the proposed algorithm and the significant role of these accelerating techniques. Furthermore, the algorithm presents its practical applicability in large-scale scenarios. The findings reveal that the proposed approach reduces the fleet size by 35 vehicles and concurrently achieves a significant 13.61 % reduction in electricity consumption, which provides a holistic solution to real-world operational challenges and offers recommendations for operators.