The field of vision-and-language navigation is moving towards more efficient and effective methods for enabling agents to interpret natural language instructions and navigate complex environments. Recent developments have focused on improving the scalability and generalizability of existing models, with a particular emphasis on addressing the challenges of limited computational resources and complex spatial and temporal reasoning. Notable advancements include the introduction of modular frameworks that decompose navigation into interpretable atomic skills, as well as the development of input-adaptive inference methods that enhance model efficiency. Additionally, end-to-end optimized policies that unify traditional two-stage frameworks have shown significant promise in improving navigation performance. Noteworthy papers include AgriVLN, which proposes a baseline model for agricultural robots and achieves state-of-the-art performance in the agricultural domain. SkillNav is also noteworthy, as it introduces a modular framework that achieves a new state-of-the-art performance on the R2R benchmark and demonstrates strong generalization to novel instruction styles and unseen environments.