Freight transport is recognized as a complex system that is affected by globalization effects, integration of different transport modes, geographically distributed operations and extended business models. Such complexity is also amplified by the need for the real-time response on the unexpected situations detected during the transportation phase (e.g. weather conditions, strikes, accidents). In a very demanding market, request for on-time cargo delivery, transport efficiency is a critical issue, and the ability for real-time detection and resolving of all possible obstacles and exceptions becomes a core competency for logistics operators. The work presented here introduces the concept of Smart Cargo, able to autonomously react to its context, find and understand alternatives, compute adaptive behavior, optimizing its own decisions. The Smart Cargo concept, implements a paradigm shift requiring taking the control of computing applications, data, and services away from some central nodes (the “cloud”) to the other logical extreme (the “fog”) of the Internet. Fog provides an intelligent connection of people, processes, data, and things, enabling Smart Cargo to go beyond the existing Intelligent Cargo concept. Fog computing is described here, as a necessary paradigm shift on distributed freight intelligence, allowing multimodal freight transport to achieve real-time situation awareness and reaction, planning with predictions and learning from the environment.