The EU has set aggressive energy targets that will drive a significant increase in the use of renewable energy sources introducing large uncertainties in the power generation as well as the uptake of plug-in hybrid electric vehicles (PHEVs), offering a new source for balancing the power grid by providing large scale distributed storage capacity. The COGENT project will develop theoretical frameworks and practical tools for applying control and optimization to the energy management of PHEVs that are active participants in the Smart Grid in an interrelated way. Innovative control strategies are required that utilize variable loads and storages to ensure optimal performance and safety of the Smart Grid based on large amounts of data. The proposed project will address this problem by developing novel and sophisticated control and machine learning techniques, that are well suited for the increasingly complex, uncertain, high-speed and distributed characteristics of the Smart Grid. The success of this challenging and highly interdisciplinary program will be enabled by the collaboration with the host institutions, the Control, Intelligent Systems and Robotics group at UC Berkeley, a center of excellence in control and the Max-Planck-Institute for Intelligent Systems, a research group world famous for its work on machine learning. The fellow researcher provides an excellent background for the proposed project, which will be supported by the outstanding expertise of the host institutions, together providing the comprehensive knowledge required for the scientific and technological goals of the program. In addition to the development of high-quality and cutting edge research, the aim of this project is to promote the young outgoing researcher by boosting her skills, expertise and international network. The project will thereby equip the researcher with essential competences and prerequisites for an academic career in Europe and for contributing to European research excellence.