In this talk we consider a nesting problem: given a container and a set of objects, we want to find the optimal location of the objects inside the container. We formulate the problem as a constrained, finite- sum minimization problem. The problem is low dimensional, however the objective function is expensive to evaluate, and its derivatives are not easily available. Moreover, because of the form of the objective function and the topology of the container, the problem is highly non-convex. To enforce the belonging of the object to the container we employed a penalty strategy. We solve the resulting unconstrained method using CBO, a particle-based method for global optimization, combined with a mini-batch strategy [1]. We present numerical results that show the effectiveness of the algorithm, and we study the performance of the method for different choices of the parameters and different penalty and sampling strategies. Nesting problems arise in several applications. We are mainly interested in the solution of problems arising in Autologous Ear Reconstruction, a surgical procedure where the external ear of patients is reconstructed using tissue from the patients themselves. The main elements of the ear are carved from a graft of rib cartilage, then used to create the framework of the new ear. The shape of the elements to be carved makes this procedure particularly complex, so several instruments have been developed to help the surgeon and partially automate the procedure. One of the main issues is to decide where, in the rib, each ear element should be carved, taking into account that the size of the rib graft should be as small as possible, and that carving along the border of the rib facilitates the procedure and increases the quality of the reconstruction. This gives rise to a nesting problem.