Privacy threat analysis and evaluation of Privacy Enhancing Technologies for the integration of 3D Scene Graphs in smart buildings

Sammanfattning: The recent advent of 3DSG in the computer vision domain has brought powerful high-level representations of 3D environments. These representations strive to mimic human perception and facilitate the extraction of meaningful insight from visual data. This research has led to framework like Hydra, which creates a real-time, persistent spatial perception system that creates a complete 3D map of their surroundings. Furthermore, smart cities are at the forefront of integrating information technology into urban life, with smart building playing an essential role. Smart buildings are equipped with IoT devices that promote their interaction to achieve a more intelligent environment. The integration of spatial perception systems like Hydra into the smart building domain can improve decision-making, autonomy, and responsiveness. However, this integration introduces a new dimension of privacy concerns among users. This arises due to Hydra’s capacity to capture sensitive information, raising valid concerns regarding the potential misuse by third parties or attackers. This necessitates a comprehensive approach that helps practitioners to systematically identify and assess potential threats and vulnerabilities to a system, application, or network. This thesis places a primary focus on addressing privacy concerns within the integration of Hydra into the smart building domain. To achieve this, this study applies the LINDDUN privacy threat methodology, which is designed to focus on privacy concerns, guiding practitioners in identifying and mitigating privacy threats. Thus, the main the goal of this thesis is to thoroughly examine and address these threats, all with the ultimate goal of protecting personal information.