Picking and placing parts is a fundamental operation involved in circuit board manufacturing and assembly plants. Pin thru-hole (PTH) components with non-standard features are incompatible with most existing high speed pick-and-place systems; this results in the use of manual operators to perform the circuit board assembly process. The inclusion of a human element in an otherwise automated manufacturing plant introduces a variety of unfavorable effects including decreased throughput rates and increased errors. In this research, I have developed an initial framework which allows for automated circuit board assembly of non-standard electrical component types. An articulated robotic manipulator with 6 degrees of freedom (DOF) was guided by precisely calibrated machine vision cameras to obtain workspace information. A model based object recognition approach was used for the identification and subsequent localization of components and their placements, which demonstrated robust and repeatable results among the different components tested. The experimental data demonstrates a decrease over the current average human throughput rate as well as reduced error frequency, with future work identified to further optimize the performance results.