Abstract This two-part paper presents two different manufacturing feature extraction approaches and the comparative studies on the two approaches. Part I will focus on presenting the optimal volume segmentation approach to feature extraction. The optimal volume segmentation approach suggests that the material to be removed from a stock to get the desired part geometry is formed by elementary volumes that can be removed in a single tool path. These elementary volumes can be grouped into a number of machinable volumes (manufacturing features). There are many different ways to group these elementary volumes, which may result in different volume removal cost, tool utilization and fixture utilization cost. A mathematical programming model for optimal selection of machinable volumes is presented in this paper. The selection of machinable volumes (feature extraction) is called optimal if the maximum amount of material can be removed in each setup. The number of setups and the cost to manufacture the component therefore are minimized too. Two powerful optimization methods, viz. Simulated Annealing and Genetic Algorithm, are used on the optimization problems.