Gears and sliding mechanical contacts are indispensable elements in power-transmission systems, where material selection, tribological behavior, thermal stability, and fault prediction govern functional life and reliability. The current article reviews three major research directions: (1) mechanical and tribological characteristics of additively manufactured polymers used in gear applications, (2) steel-on-steel tribological systems and the integration of artificial intelligence (AI) and physics-informed machine learning (PIML), and (3) advancements in gear dynamics of combined material gears, and fault detection methods. The additive manufacturing process, fused deposition modeling (FDM) especially, has facilitated fast production of complex-shaped gears using polymers; yet these gears lack the capability to handle loads owing to their anisotropic property, low stiffness, thermal sensitivity, and surface roughness. However, gears made of polymers offers light weight and self-lubrication. The steel gears offer far greater loading capability, stiffness, durability under fatigue loading, and thermal stability; yet their noise level, weight, and machinability can be prohibitive factors at times. Emerging combined material designs attempt to merge the benefits of both materials by improving heat dissipation and wear resistance. Meanwhile, tribological research—especially involving steel systems—has advanced through machine learning and physics-informed neural networks (PINN) architectures that enhance friction and wear prediction. Conventional vibration analysis techniques for gear diagnosis have been expanded to artificial intelligence-based systems with the capability of identifying variances in worn-out mechanisms, simulating crack growth patterns, and solving equations for distributed damage models.