This thesis focuses on the management of disruption in freight transport. The management of disruption is of importance to achieve planned efficiency in the transport system by mitigating or avoiding the impacts of disruptions, such as late arrivals of deliveries. The transport system is influenced by the ongoing development of information and communication technologies, which makes real-time information related to the transport operations available. This information has been shown to be useful for disruption management and has created possibilities for a shift towards more autonomous transport systems. This thesis investigates the real-time disruption management, as the recovery phase of disruption at the operational level, using relevant information after a disruption has occurred. The studied literature has mainly considered this phase to be reactive and focused on recovery strategies as proactive actions before disruptions occur. This thesis considers proactive recovery actions as made after a disruption but before it impacts the transport chain. These kinds of recovery actions provide less impact from disruptions on the freight transport system than reactive recovery actions made after the impact has occurred do. The purpose of this research is to investigate real-time disruption management in freight transport, in order to generate possibilities for proactive recovery actions. Two cases of real-time disruption management have been investigated in this thesis, in three different studies. Each study examined different aspects regarding detection of disruptions, of what is detected, how it is detected and where in the transport system it is detected, influence on the initiation of real-time disruption management. The results from the performed studies point towards the importance of that detecting different objects of a disruption, which is further influenced by how and where in the system the detection is made. Furthermore, these insights into the detection phase are connected to the other phases for real-time disruption management, prediction and action, in order to state the possibilities of generating proactive recovery actions. In contrast to the developed literature of strategic recovery strategies, this thesis establishes a detailed description of the viewpoint of real-time management of disruptions. As the identified objects for detection in this research are shown to be represented by different information, it is valuable for the development of disruption management to match future autonomous parts of the transport system and develop decision support systems accordingly. Furthermore, the research contributes with two dimensions in which recovery actions can be viewed as proactive. This is generated either with real-time disruption management performed after impact but before impact on the transport system or after the impact on transport system but before impact on upcoming operations. The practical contribution includes concepts revolving around real-time disruption management, which can be used for an outline for needed information in order to generate possibilities for proactive recovery actions.