The damage condition of paved roads in Indonesia will increase as the number of vehicles passing and erratic weather conditions. One way to reduce road damage is to increase the quality of the asphalt mixture. To increase the quality of asphalt can be done by adding additional materials such as plastic waste into the asphalt mixture. In this study, simulated use of neural networks to determine the optimal value and analysis of the influence of plastic waste polypropylene and polyethylene terephthalate to the mechanical and thermal properties of the asphalt. Based on the design of ANN, obtained the output accuracy of 81% for PP and 91% for PET. Based on the results of the test and simulation, the addition of PP and PET plastic waste can reduce the penetration value and the softening point that affects the asphalt resistance to temperature changes, as well as increase the value of stability that affects the maximum load acceptable to the asphalt. The best composition of AC-WC modification that has been by the specifications of Bina Marga for PP is 1% PP and 5.6% optimum asphalt content, while for PET is 5.95% PET and 6% optimum asphalt content.