Many optimization problems in science and engineering fields are large-scale and heterogeneous, and have many conflicting objectives of different nature, many decision variables and multiple sources of uncertainty. It is very difficult for traditional optimization methods to deal with these BIG optimization problems. This proposal is to hybridize techniques and theory from evolutionary computation, machine learning, optimization, and use modern parallel platforms for developing efficient big multiobjective optimization solvers. We will adopt an evolutionary decomposition framework to decompose a multiobjective problem into a number of single-objective or simple multiobjective subproblems and then solve them in a collaborative manner. We will study how to use landscape analysis to identify major problem features that make a big multiobjective problem difficult, and investigate how to do configuration and adaption in the algorithmic framework. We will study how to use surrogate meta-models and how to distribute the computational resources in distributed computing environments. Different parallelism mechanisms will be studied in this project.