Google crowd sources its traffic data. Smartphones with Google Maps installed send anonymous bits of data back to Google describing how fast they are moving. These bits of data are combined across multiple users, and stored over time. This creates a historic database of traffic, and a model of present time traffic.
This feature launched Google Traffic in the United States in 2009, and probably started gathering data for the rest of the world. It continued to expand across more cities ever since. Google bought Waze in 2013, which added even more data and included things such as information on accidents and road blocks. However, it has not launched Google Traffic in Egypt as claimed, as there has been no announcement on its official blog as is customary.
Rather, this feature is likely a byproduct of a new product Google launched 2 weeks ago, when it released a new API to predict Future Travel Times. An API is a piece of software that developers use to design products; imagine a foundation to be built upon. Furthermore, it is only accessible via a workaround as described by AndroidHeadlines and AndroidPolice.
It is important to understand the limitations of these traffic predictions. With no information on the quantity or quality of the traffic data or an official launch -and thus endorsement as use-ready from Google- the predictions supplied are likely to be inaccurate. Predictions rely on algorithms, mathematical formulas used to create predictions. These are global ones that have not been attuned to the local context of Egypt and are thus relatively unreliable.
In short, proceed with caution.
 “Predicting Future Travel Times with the Google Maps APIs.” Official Google for Work Blog. Accessed November 23, 2015. http://googleforwork.blogspot.com/2015/11/predicting-future-travel-times-with-the-Google-Maps-APIs.html.