Currently there are a large number of metrics to measure the quality of software architectural design, however, there are still missing metrics that are not defined in the literature, such as metrics to measure the protection of methods associated with the "Template Method" design pattern. In this work, a research applied to technological development was carried out based on the theory of measurement as a means of validation and combination of empirical conditions with formal conditions. The main result of the research was a set of four metrics that allow to evaluate the internal quality of software architectures in order to facilitate their maintenance and useful life time, as well as their reuse. Three of them measure the degree of modular protection in template methods (PMMP), variant functions (PMFV) and invariant functions (PMFI). The fourth one measures the degree of total protection that the Template Method design pattern has in a software architecture (PTTM). Finally, the metrics were implemented in a system to automate the calculation applied to three test cases where the validation was given by comparing the expected results with the obtained values. Keywords: Template Method, functional protection, variant functions, invariant functions, indirect coupling.

One of the bad decisions made by a developer is the lack of a good encapsulation in the design pattern "Template Method", when this happens, "Code Smell" appears, which is defined as "symptoms of poor design and implementation choices", these symptoms will be reflected in a high degree of dependency between classes of a program (indirect coupling), thus shortening the lifetime of the software.

Municipal Solid Waste (MSW) is a common problem in all cities worldwide; it is expected to increase to 3400 billion tons by 2050. In Mexico, an average of 108,146 tons of MSW are generated daily. Artificial Intelligence (AI) is a computer tool that allows the development of systems that facilitate the recycling process. However, most AI programs focus on classifying paper, plastic, glass and metal; therefore, wood and textile waste have received little attention. Using texture techniques such as Local Binary Pattern (LBP), Gray-Level Co-occurrence Matrix (GLCM), Histogram of Oriented Gradients (HOG), Canny/Sobel edge detection, Fractal Dimension (FD), feature values were extracted and integrated from 4396 images belonging to wood and textile categories. Using the Random Forest Importance method, the most significant features were selected to train three Machine Learning (ML) algorithms. Multilayer Perceptron (MLP) achieved the best performance in accuracy with 96.70%, followed by Random Forest (RF) at 95.45% and Support Vector Machine (SVM) with 95.22%. The implementation of these comparisons will serve as a basis for the development of new technological tools with low computational cost that carry out a proper waste separation.