Control Networks Group

The decreasing costs of embedded computing and communication technologies are pushing several of today’s engineering systems toward increased levels of autonomy. Highly interconnected networks of physical devices and embedded computers, exhibiting continuous (due to the physics) and discrete (due to the computation) behavior, that is, Cyber-Physical Systems (CPS), are becoming a substantial part of our life. These systems range from transportation networks, to smart buildings, to the power grid. While the potential for improvement is enormous, increased levels of autonomy also bring about critical problems. Challenges include ensuring a correct and safe functioning despite the large number of physical devices, the intrinsically hybrid continuous/discrete nature of the system, limited information, and the presence of human operators. How can one design these systems so they are provably correct despite these challenges? How does one formally verify their behavior? Del Vecchio’s group is addressing these questions by merging tools from control theory and computer science. While the first has historically been used to design physical devices, the second has been employed to design and analyze computer programs. Specifically, Del Vecchio formally models a Cyber-Physical System as a hybrid automaton with imperfect information and mathematically formulates the design/verification questions as a safety control problem, seeking computationally efficient solutions.

Modularity is the property that allows to predict the behavior of a complicated system from that of its components, guaranteeing that the properties of individual components do not change after interconnection. Researchers in the systems biology community have hypothesized that modularity may be a property of biological systems, proposing functional modules as a critical level of biological organization. This view has profound implications on evolution and also implies that biology, just like engineering, can be understood in a hierarchical fashion. However, modularity is still subject of intense debate and remains one of the most vexing questions in systems biology. At the same time, modularity is critical for the field of synthetic biology, in which researchers are pursuing a bottom-up approach to design. Del Vecchio's group addresses this question, by employing and extending the theory of nonlinear control systems, with the ultimate goal of establishing fundamental theory for modular biomolecular network analysis and design. This theory seeks to (a) shed light on the extent to which modularity is a principle of biological organization, (b) provide techniques to analyze networks modularly, and (c) empower synthetic biology of a number of concrete design tools to enforce modular composition, which will move the field forward.