Automatic optimum design based on high-level design requirements, such as the quality of the expected system, is a new and interesting field of study in the present era. In that regard, in the present thesis a new algorithm is proposed and developed. This algorithm which is named Vanguard Frontier is partially iired by two known ideas: Imperialistic Completion Algorithm which is an optimization algorithm operating in the data space and the other one which is known as Genetic Programming which works in the topology space. The new algorithm which is abbreviated as VF, is aimed to automatically propose 2D multidisciplinary monolithic designs, focusing on their topology. Monolithic design, in comparison with hinged or other multi-part designs, has many advantages such as easier manufacturing process, lower assembly costs, lower kinematic noise, no wearing phenomena, high preciseness and the ability of creating non-convectional schemes. Such innovative designs could be very useful in many new fields such as energy harvesters, vibrational transducers, surgical instruments, Micro Electro-Mechanical Systems (MEMS) etc. In order to show the abilities of the developed method and its corresponding software, two benchmark problems are firstly solved. The first one is a beam, subjected to pure bending load and the second is the same as the first, with a different loading, which is a pure torsion. Both are intended to have a cross section with the highest stiffness to weight ratio. The final problem includes designing an energy harvester inside a limited area with the lowest possible weight. The results are then compared to an already optimized solution which have proved to have a better quality. Keywords: Vanguard Frontier Algorithm, optimum design, monolithic systems, imperialistic competition algorithm, genetic programming, energy harvesting