This project aims to develop integrated formalisms for self-reconfigurable robot modeling and multi-manual manipulation that allow the application of such platforms to inspection tasks in hazardous and hard-to-access environments, such as nuclear power plants and offshore platforms. The modeling should be general enough to allow the online inclusion of new modules into the system, and the multi-manual manipulation should take advantage of the modularity and versatility of the robot. Addressing those problems involves modeling, control, and planning techniques. Indeed, as the number of modules and the diversity of their characteristics increase, so does the computational complexity of obtaining the dynamic model of the robot. Decentralized approaches can delegate this process to its modules. However, rather than having predefined solutions to a given set of modules, which would limit the system's versatility, one challenge is for the robot to update its model when including new and possibly unknown modules. The control of such a complex robot also raises important questions. For instance, suppose the robot uses two or more manipulators to carry a heavy object. The internal stresses in the manipulated object, generated by the use of multiple manipulators, require the use of a force control scheme. If a given module performs this task, can this control scheme also be decentralized and, thus, leave the other modules free to perform different tasks? If those manipulators are made of two or more modules, how can they cooperate? Addressing those questions leads, in turn, to the self-reconfiguration planning problem: How can the robot decide, in real-time and preferably at the task level, which modules should engage in a given task? Can it optimize the number of available modules to deal with unpredicted situations, such as balance loss due to collision with the environment or uneven terrain? The development of this project is crucial to the realization of any self-reconfigurable robot application. Developing this system will significantly benefit operations in hard-to-access locations, such as rescue missions or inspections in hazardous sites. Furthermore, applying these robotic platforms to decommission nuclear power plants will have a significant economic and public health impact in several countries, including the United Kingdom and European nations.