Banca de QUALIFICAÇÃO: Lara Monalisa Alves dos Santos

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
STUDENT : Lara Monalisa Alves dos Santos
DATE: 07/03/2023
TIME: 14:30
LOCAL: Videoconferência
TITLE:

Building inspection in flat roof: use of digital image acquisition technologies and automated processes

KEY WORDS:

Maintenance; degradation; flat roofs; deep learning drones; equipments.

PAGES: 145
BIG AREA: Ciências Sociais Aplicadas
AREA: Arquitetura e Urbanismo
SUMMARY:

Building inspection processes must be carried out periodically to diagnose the condition and plan maintenance. Scheduled building inspections require technical knowledge and can be based on sensory analysis. In roofing systems, building inspections may involve difficult access, security risks due to work at heights, and time spent to carry out the inspection. As part of the building envelope, the roof must meet the functional requirements related to water tightness, the thermal performance of the internal environment, and the durability of the building. In roofing systems with flat roofs, the intensive use of the roof is recurrent to accommodate parts of the building installations and different equipment. In this context, on-site surveys and the identification of pathological manifestations for repairs and replacements of elements or components of the flat roof, or equipment placed on top of it, present a situation of complexity to be researched. Through a systematic review of the literature, a gap was observed regarding the integrated approach to inspecting the flat roof and assessing their state of conservation, considering the types of equipment present, their installation processes, and the interface with the waterproofing layer, slope, and drainage. That said, this research questions how the use of digital technologies and automated processes can contribute to the inspection processes of flat roofs. Still, we seek to investigate in an integrated way how the presence of equipment and its installation processes affect the required performance and conservation status and interfere with maintenance actions. In this sense, this research aims to contribute to the improvement of inspections and building maintenance in flat roof systems with the use of digital image acquisition technologies and automated processes. As for the design of the research method, descriptive research will be used to characterize the use of digital technologies and automated processes, the systems in focus, and the reference methods of inspection and evaluation relevant to the study, in addition to exploratory research in field surveys and experiments for the application of these technologies. The proposed method associates several tools for visual inspection through images, the construction of a database, and an analytical structure with a set of criteria. The sample selected for the exploratory studies consists of the nine flat roofs of the building set of the Headquarters of the Brazilian Army and the flat roof of the Central Library of the University of Brasília, located in Brasília-DF. The partial results obtained in the studies already carried out for this research show satisfactory performance in the application of convolutional neural networks for the detection of condenser-type equipment present in the roof and the areas affected by the pooling. With this thesis, it is expected to contribute to a methodological approach within the scope of the integrated vision of maintenance management and
intervention priority, as a strategy to speed up preventive and corrective maintenance actions.

BANKING MEMBERS:
Presidente - 1851456 - VANDA ALICE GARCIA ZANONI
Interna - 1438937 - JOARA CRONEMBERGER RIBEIRO SILVA
Externo ao Programa - 2712272 - CLAUDIO HENRIQUE DE ALMEIDA FEITOSA PEREIRA
Externa à Instituição - Cristiane Machado Parisi Jonov - UFMG
Externa à Instituição - DAYANA BASTOS COSTA - UFBA