Abstract
<title>Abstract</title> <p>Urban rail transit systems are the backbone of city mobility, yet remain vulnerable to disruptions that can affect hundreds of thousands of passengers daily. Bus bridging provides temporary supplementation for rail services during planned and unplanned disruptions. This paper systematically reviews 97 most related studies published between 1995 and 2026, with three primary objectives: (i) to clarify key gaps in the literature; (ii) to synthesize findings across methodological approaches; and (iii) to guide future research directions. It categorizes existing studies into three operational contexts: nominal operations (planned integration), disruption-response operations (reactive deployment) and robust operations (preventive reinforcement). The review examines current models, solution methods, and real-world applications in these areas. It traces the shift from early, basic optimization and evacuation models to more advanced approaches that use hybrid methods. However, there are still major challenges, including computational complexity, limited real-time adaptability, and poor integration across transport modes. The findings point to several ways to make bus bridging a proactive part of resilient transit planning. These include using real-time data, creating adaptive optimization tools with stochastic models, considering passenger behavior, and using simulation-based decision support. Adopting frameworks such as the preconditioned inexact stochastic alternating direction method of multipliers and its variants, which are designed for complex, distributed optimization and provide convergence guarantees under mild assumptions, will help manage disruptions in real time. These improvements can transform bus bridging from reactive emergency into a smart, fair, and essential part of resilient urban transport.</p>