The first method to automatic identify legacy structure plants (copper wire lines and poles) in Plain Old Telephone Systems (POTs) (but not limited to). The process makes use of legacy database information, geographical characteristics, and cost assessment, combining machine learning procedures with sophisticated mathematical optimization to automatically identify removable infrastructure cost-efficiently. The automation in this process reduces the human intervention, mainly in legacy and incomplete/incorrect data, speeding up the assessment and optimizing the deployment of field technicians, aiming to save in operations costs.