XVIII. MİMARLIKTA SAYISAL TASARIM ULUSAL SEMPOZYUMU
A Neuro Fuzzy Logic Based Artificial Intelligence Course Application in Architecture Education
Yayıncı:
Balıkesir Üniversitesi Yayınları
Artificial intelligence has recently become a technology that users from all walks of life can benefit from in different ways through various applications. Artificial intelligence is able to organise the data compiled by scanning large alpha-numerical and visual data stores in a way to meet the specific needs of users from different fields of study, express them with various representation methods and contribute to the production and informed decisions of users. Many methods based on artificial intelligence, which have been used in engineering fields in the past, have increased their use in creative fields such as architecture. The production of visuals that relate to material, light, form and context with AI-supported visualisation techniques accelerates the creation of the design concept and allows the production of alternatives that would take a long time in other numerical methods. Artificial intelligence has long contributed to the generation of geometric solutions in problems involving multi-criteria and complex parameters by quickly considering design inputs and constraints holistically. Methods such as artificial neural networks, fuzzy logic, neuro fuzzy logic and ANFIS (Adaptive-Network-Based Fuzzy Inference System), which are among the applications of artificial intelligence in the design phase, can provide significant benefits for the evaluation of realistic results before the building is implemented; thus, it is possible for the design outputs to provide more sustainability and comfort conditions. This research shares the educational experience gained in the bachelor elective course ‘Architecture and Artificial Intelligence’ in the curriculum of Antalya Bilim University, Faculty of Fine Arts and Architecture, Department of Architecture. The use of artificial intelligence tools in relation to skills such as imagination and creativity required for design has many potentials and challenges in terms of pedagogy. It is expected that this type of educational practices, which are currently experimental, can provide significant benefits with more effective content and teaching methods in the architecture curriculum in the coming years. According to the data obtained from the questionnaire applied to the architecture students, it was evaluated that the elective course selected for the 4th year would be more beneficial if it was selected in the fall or spring semesters of the 2nd year. It has been determined that the application in the studio courses they will take after taking this course can be the period in which students can analyse the effect of design development, design decision making, comprehending the parameters that will affect energy or comfort in design. In this study, the approach of creating architectural solutions from numerical data with neuro fuzzy logic has been tested for the first time on architecture department students. As a result of the studies, the effect of different inputs on outputs was analysed by making analyses such as the relationship of indoor air quality with the number of users in the space, the effect of the facade shading element on the window wall size on the facade and the effect of the lighting elements on the direction of the building by determining more than one factor that can affect the design with classworks and assignments.
