The accelerating deterioration of road infrastructure, exacerbated by increasing traffic loads and climate variability, necessitates sustainable and intelligent rehabilitation strategies. Simultaneously, the environmental footprint of conventional bitumen underscores the need for eco-friendly alternatives. This study explores an integrated approach to sustainable pavement rehabilitation by combining waste-modified asphalt mixtures with an artificial intelligence (AI)-driven road defect management system. VG-30 bitumen was modified using lignin and waste engine oil (WEO), both individually and in hybrid formulations, to evaluate their effects on physical, thermal, and mechanical properties. Lignin improved stiffness, softening point, and rutting resistance, whereas WEO enhanced ductility and low-temperature flexibility. Hybrid mixes such as L20EO6 and L30EO9 demonstrated a balanced response, mitigating the brittleness of lignin and the excessive softening from WEO. All modified binders satisfied storage stability and thermal safety requirements. The study also developed a three-tier AI framework for automated pavement inspection and maintenance prioritization. Using a curated data set of manually annotated images, a YOLOv7 model was employed for real-time detection of potholes, cracks, and rutting, whereas U-Net segmentation quantified defect severity based on area and geometry. The models achieved high accuracy, with YOLOv7 reaching 96.8 % mAPat0.5 and U-Net providing an average Intersection over Union of 0.85. The results demonstrate that hybrid lignin–WEO-modified asphalt mixtures can offer both environmental advantages and improved mechanical performance, whereas the AI-based framework enhances efficiency in defect detection, classification, and maintenance planning. The integration of sustainable materials with intelligent defect prioritization presents a feasible solution for cost-effective and environmentally conscious pavement management in resource-constrained regions.