Networks are a cornerstone of modern digital life, and their performance depends heavily on the ability to manage dynamic traffic patterns. As we move toward future intelligent networks, there is a growing need for systems that can adapt to diverse traffic conditions without relying on static mechanisms or human intervention. In recent years, artificial intelligence (AI) techniques have proven effective in making intelligent decisions and enabling networks to respond dynamically to unforeseen conditions. Packet scheduling plays a key role in controlling network performance, but traditional scheduling methods often lack the flexibility to adapt to changing traffic patterns. To address this, recent research has explored the integration of Traffic Intensity, or Erlang-based packet scheduling, which incorporates network dimensioning into scheduling decisions. The Erlang distribution offers advantages in accurately modeling both spatial and temporal aspects of user traffic. Recent advancements have further focused on the temporal characteristics of traffic, leading to significant improvements in packet scheduling performance. This article proposes a Spatio-Temporal Traffic Intensity (STTI) model for real-time packet scheduling decisions. Early experimental results demonstrate that the STTI model more effectively captures real-world traffic patterns and provides a more autonomous and intelligent approach to scheduling in congested networks, improving overall network performance.