The escalating challenge of plastic waste and the demand for sustainable infrastructure have prompted exploration into innovative construction materials. This study assessed the feasibility of using polyethylene terephthalate (PET) waste as a partial replacement for coarse aggregate in permeable concrete. An experimental program was conducted, examining constituent materials and key concrete properties such as workability, permeability, water absorption, and compressive strength. Due to a limited dataset, a novel data augmentation approach using controlled random noise in MATLAB expanded the original 32 data points to 352. This enriched dataset was used to train predictive models: an Artificial Neural Network (ANN) and an Adaptive Neuro Fuzzy Inference System (ANFIS). Results showed an inverse relationship between PET content and compressive strength. The 28-day strength dropped from 23.1 N/mm² in the control mix to 8.7 N/mm² at 20 percent PET, representing a 62.3 percent reduction. Both predictive models performed exceptionally well, with both ANFIS and ANN achieving nearly identical and high predictive accuracy (R² = 0.9990 and 0.9988, respectively). The study concludes that up to 10 percent PET waste replacement yields permeable concrete with acceptable strength for low-load applications, such as pedestrian walkways and landscaping.