year: 2021

Limitations and Review of Geometric Deep Learning Algorithms for Monocular 3D Reconstruction in Architecture

This paper aims to test algorithms for 3D reconstruction from a single image specifically for building envelopes. This research shows the current limitations of these approaches when applied to classes outside of the initial distribution. We tested solutions with differentiable rendering, implicit functions, and other end–to–end geometric deep learning approaches. We recognize the importance of gener-ating a 3D reconstruction from a single image for many different industries, not only for Architecture, Engineering, and Construction (AEC) industry but also for robotics, autonomous driving, gaming, virtual and augmented reality, drone delivery, 3D authoring, improving 2D recognition and many others. Henceforth, engineers and computer scientists could benefit, not only from having the 3D representations but also from the Building Information Model (BIM) at their disposal. With further development of these algorithms it could be possible to access specific properties such as thermal, physical, maintenance, cost, and other parameters embedded in the class.

AR&AI in the Didactics of the Representation Disciplines

The class of Digital Representation Techniques is an optional class for architecture and engineering students at the Politecnico di Torino. For the second year, the program of the class, focused on the the-oretical and applicative panorama of technologies for digital drawing, integrates an introduction to the most recent developments in augmented reality (AR) and artificial intelligence (AI), two technological fields that are deeply influencing the interpretation of physical assets through digital models. The class requires students to produce different deliverables, including a research essay, graphic representation documents (plates or brochures), a video proving animation and editing skills, and a prototype AR ap-plication. The proposed contribution presents methods, objectives, and some of the results of the class, highlighting the interaction between the different digital representation techniques and deepening the critical aspects related to the introduction of new AR and AI technologies.

Forms in Space. AR Experiences for Geometries of Architectural Form

Learning through the direct experimentation of models, in their variety of manifestations and hybrid-izations that we know today, is undoubtedly a very powerful aid in the acquisition of knowledge. Spe-cifically on architectural form, this aid is even more important, because it helps not only to understand the form of existing architecture, but even more to imagine and design new ones. The research here presented, focusing on this double objective, cognitive and creative, proposes and experiments new ways of integrating and interacting with heterogeneous models – both physical and virtual – conceived for a scenario of musealization of the architectural form. A place where the user interacts and experi-ences the properties and peculiarities of form, in perceptive continuity between real and virtual space.

The Drawn Space for Inclusion and Communicating Space

In the last twenty–five years, some studies have been analyzing the virtual reality potential for special education. They show the virtual environments as a valid communicative medium and a safe space where ASD people can experience new situations without limits of ‘in vivo’ experiences. Often, prob-lems with space can complicate many aspects of everyday life. Referring to Hermelin and O’Connor’s studies, the difficulties involving the autistic clinical frame are connected especially to perception deficiency, therefore the VR can become valid support, for people with ASD, improving relationships with space, with ourselves and with others. My study tries to provide a guideline tool for a human–centered VR design.

Genetic Algorithms for Polycentric Curves Interpretation

This research work into Evolutionary Computing field aims at improving a dataset of algorithmic gen-erative definitions able to return an optimized ‘semi–ideal’ curve that best fits a generic reality–based profile, starting from some of its points. This paper shows GAs_Genetic Algorithms applications especially with regards to study, interpretation and definition of generic polycentric curves. Current VPL tools (Galapagos–Rhino, McNeel) allow to test Evolutionary Theories for problem solving and decision makingin architectural research field. According to a human driven approach, an operator defines GENOME, FUNCTION and FITNESS to drive the Evolutionary Solver towards optimized solutions. Some case studies from Historical/Existent Architectural Heritage are used to show how GAs can simplify the digitalization process and big data interpretation.

Visual Languages: On–Board Communication as a Perception of Customercaring

The year 2020 should have been one of confirmation of the continuous growth in the “cruise tour-ism sector”: new ships for delivery, new constructions, and the design trend of the gigantism of these means of transport. If, at first, ensuring a correct perception of orientation on board these new ships were a complex challenge and not yet fully resolved, today with the outcomes of the Covid pandem-ic it appears even more and more complex.The logistical communication of the past now seems to be no longer sufficient: the collapse of cruise bookings has introduced serious problems in the sector which must not only rethink new models of internal organization on board, but also reconsider how to regain the trust of potential passengers, frightened at the idea of getting on a ship.This research introduces some reflections and proposals shared with industry operators and aimed at the reformulation of approaches and languages in order to bring potential customers closer to the Cruise Experience.

V.A.I. Reality. A Holistic Approach for Industrial Heritage Enhancement

The post–industrial heritage, characterized by heterogeneous, tangible and intangible factors, requires digital tools and a holistic approach to undertake the most appropriate enhancement process. Cur-rent virtual realities (Vr, Ar, Ir) allow the modeling of physical environments and the management and virtualization of a large and varied amount of data, thus helping to better understand the complex-ity of the real phenomenon. The paper proposes a method, named D.V.M.R. (acronym for Design, Virtualization, Modeling, Reproduction), which in four temporally consequential phases builds a tool capable of providing territorial, environmental, architectural and historical information of a case study. The method was applied for the design of a reuse of the brick factory, known as ex Sieci and located in Scauri in southern Lazio, owned by the Municipality of Minturno. The factory, which looks like a majestic cathedral on the sea, had in the past and still has a significant centrality in the social life of the local inhabitants.