Ing. arch. Lukáš Kurilla, Ph.D.

Ing. arch. Lukáš Kurilla, Ph.D., Ing. arch. Vasilisa Supranovich

The latest trends in computer graphics show that developers, in pursuit of the perfect image, are increasingly creating graphics that are difficult for the user to perceive. The struggle for impressive graphics has begun to raise questions among users more and more often, as graphic characteristics are infinitely approaching the level of hyper-realism, which in turn causes the phenomenon of underrealism and broken immersion in the virtual space. In connection with this, the question arises - Is there sufficient abstraction of the virtual reality object and space that has the same effect on user immersion as graphically more complex variants? This paper is mainly divided to three parts, in first part, before proceeding to the analysis of existing research and examples from various fields of computer graphics, a few necessary terms are explained at the beginning of the article. First of all, the differences between the feeling of immersion and the present by the user in virtual reality are determined, since the level of detail of the scene has a direct impact on the user’s perception of the virtual space. The second necessary concept is the uncanny valley phenomenon. This phenomenon determines that the realism of the avatar graphics can produce a negative impression on the user. In this work, avatars are considered as components of virtual scenes; therefore, their detailing is important at the level of visualization of the surrounding spaces in virtual reality. After introducing the necessary definitions, this article will discuss and analyze examples of different types of graphics of virtual scenes, such as architectural visualization, computer game graphics and animated movies. Each of these fields and industries in its own way affect the development and perception of computer graphics by the user, so it is important to pay attention to each one of them. The last part of this paper will focus mainly on the methodology of a future experiment, which is intended to be used to complete the main objective – find the level of detail and abstraction of virtual reality objects sufficient to achieve the necessary user involvement and sense of presence in a given space. It is important to note that research of a sufficient level of detail of virtual reality objects will solve several problems at the same time: it will help the creators of games and virtual environments to use hardware more efficiently and create less demanding products that will have the same impact on users as graphically more complex variants and will be more accessible to a wider group of users.

Detail

Ing. arch. Lukáš Kurilla, Ph.D.

The facility layout problem is one of the fundamental production system management problems. It has a significant impact on overall system efficiency. This paper introduces a new facility layout problem that allows for choosing from multiple variants of each facility. The need for choosing the most suitable selection from the facility variants while at the same time optimizing other layout quality indicators represents a new optimization challenge. We build on our previous work where single- and multi-objective evolutionary algorithms using indirect representation were proposed to solve the facility layout problem. Here, the evolutionary algorithms are adapted for the problem of facility variants, including the new solution representation and variation operators. Additionally, a cooling schedule, whose role is to control the exploration/exploitation ratio during the course of the optimization process, is proposed. It was inspired by the cooling schedule used in the simulated annealing technique. The extended evolutionary algorithms have been experimentally evaluated on two data sets, with and without the alternative variants of facilities. The obtained results demonstrate the capability of the extended evolutionary algorithms to solve the newly formulated facility layout problem efficiently. It also shows that the cooling schedule improves the convergence of the algorithms.

Detail

prof. Dr. Henri Hubertus Achten, Ing. arch. Lukáš Kurilla, Ph.D., Ing. arch. Jiří Vele

The traditional method of data preparation for 3D printing, known as planar slicing, involves slicing the 3D model into horizontal layers and printing them gradually. This is the simplest option with the main parameter being layer height. However, this method has several limitations, including poor surface finish with stair-stepping contours of layers at steep angles and a necessity to print additional support structure for overhangs. Nonplanar slicing is a newer method that involves slicing the 3D model into non-horizontal layers with varying layer heights. This technique is mostly explored in plastics, with observed improvements in buildability, surface finish and reduction of cracking alongside the layers. In construction scale, non-planar printing is used primarily for achieving unique surface finish, or for printing on an uneven base. Its potential for improving buildability is still yet to be properly tested. This paper examines how non-planar layers can be derived with a help of force flow lines and how it affects the buildability. While printing overhangs, shear force can ultimately break the interlayer bond and layers can start deforming and sliding on top of each other, resulting in buckling, or even collapse. By guiding these forces into the bulk of layers instead of interlayer bonds, printing capabilities can be improved. Goal of this paper is to present how buildability of overhangs can be improved using non-planar slicing. Non-planar layers are derived from force flow line simulations done in Karamba3D. For printing we use clay, as a fast and simple prototyping method with the aim to later utilise our findings into concrete printing.

Detail

Ing. arch. Lukáš Kurilla, Ph.D., Ing. arch. Adam Novotník, Ing. arch. Vasilisa Supranovich

Learning in virtual reality (VR) is always happening in a certain digital space. Inspired by memory techniques such as Method of Loci AKA Memory Palace, we propose to use a VR 3D space as a form of Memory Palace which would enhance the learning process by attaching new knowledge to spatial experience. Aim is to create multi-sensory experiences and help people to form deeper memory traces of learned topics. To achieve desired efficiency we need easily remembered space, which would help to create rich episodic memory. It is also important for the environment not to disturb and attract too much attention from the learning topic. To find this thin line, we tested memorability of VR space through experiments with 9 different environments each deferring in the level of abstraction and detail of the scenes. Subjects walked through the environment in VR and were instructed to perform various memory tasks. Results show that an environment with a middle level of abstraction and detail was easiest to remember, since it achieved a certain level of newness while being well comprehensible.

Detail

Ing. arch. Lukáš Kurilla, Ph.D., Ing. arch. Adam Novotník

Learning in virtual reality (VR) is always happening in a certain digital space. Inspired by memory techniques such as Method of Loci AKA Memory Palace, we propose to use a VR 3D space as a form of Memory Palace which would enhance the learning process by attaching new knowledge to spatial experience. Aim is to create multi-sensory experiences and help people to form deeper memory traces of learned topics. To achieve desired efficiency we need easily remembered space, which would help to create rich episodic memory. It is also important for the environment not to disturb and attract too much attention from the learning topic. To find this thin line, we tested memorability of VR space through experiments with 9 different environments each deferring in the level of abstraction and detail of the scenes. Subjects walked through the environment in VR and were instructed to perform various memory tasks. Results show that an environment with a middle level of abstraction and detail was easiest to remember, since it achieved a certain level of newness while being well comprehensible.

Detail

Ing. arch. Lukáš Kurilla, Ph.D., Ing. arch. Šimon Prokop

Due to their complex geometry, it is challenging to assess wind effects on the freeform, double-curved building facades. The traditional building code EN 1991-1-4 (730035) only accounts for basic shapes such as cubes, spheres, and cylinders. Moreover, even though wind tunnel measurements are considered to be more precise than other methods, they are still limited by the number of measurement points that can be taken. This limitation, combined with the time and resources required for the analysis, can limit the ability to fully capture detailed wind effects on the whole complex freeform shape of the building. In this study, we propose the use of neural network models trained to predict wind pressure on complex double-curved facades. The neural network is a powerful data- driven machine learning technique that can, in theory, learn an approximation of any function from data, making it well-suited for this application. Our approach was empirically evaluated using a set of 31 points measured in the wind tunnel on a 3D printed model in 1:300 scale of the real architectural design of a concert hall in Ostrava. The results of this evaluation demonstrate the effectiveness of our neural network method in estimating wind pressures on complex freeform facades.

Detail

Ing. arch. Lukáš Kurilla, Ph.D., Ing. arch. Lucia Cyprianová

This review paper summarises the current state of digital simulation methods used to analyse human behaviour in physical environments. It aims to identify why digital simulations of human behaviour are not yet commonly used by architects and urban planners and to provide a systematic review of available methods, software tools, their capabilities and limitations. It covers physical models, Space Syntax, virtual reality and dynamic simulation methods for pedestrian and occupant models. There are a number of most common decision-making processes for pedestrian and occupant models. Rule-based process involves predetermined rules or heuristics. Markov Chain uses probabilities and the current state to determine state transition. Goal Oriented Action Planning chains predefined activities to achieve a specified goal. In Narrative-based modelling a narrative governs agent behaviour. In Game-theoretic approach interactions are modelled as a game. Reinforcement Learning involves learning through rewards or punishments and Artificial Neural Networks learn from data and make predictions based on that learning. Although Artificial Neural Networks come with many benefits, such as high flexibility and the ability to handle uncertain and incomplete data, it is a relatively new field of research and it requires significant computational resources. Each process has its own strengths and limitations. Multiple algorithms or decision-making processes can be combined to achieve the desired simulation properties. There are multiple digital simulation software tools for pedestrian and occupant simulations. Some of them, such as SUMO, Aimsun, and MATSim, are optimised for transportation simulations but are not suitable to become auxiliary tools in the architectural design process. PathFinder is a software for simulating evacuation scenarios. Anylogic is a flexible user-friendly tool suitable for a wider range of applications. Tools like IES and CitySim are used for building performance optimisation. A game engine Unity is a suitable option for creating interactive and visually appealing models, however it requires programming knowledge. A simulation in a 3D environment is more complex and requires more computational resources, but it can bring more realistic results compared to simulations in 2D. The paper concludes with discussion about the reasons why digital simulation has yet not become a common tool in architectural practice.

Detail

doc. Ing. arch. Jana Zdráhalová, Ph.D., Ing. arch. Lukáš Kurilla, Ph.D.

Tento text zkoumá možnosti využití umělé inteligence pro interpretaci aktivit ve veřejném prostoru měst. Aplikací neuronové sítě na kamerové záznamy, které po čtyři dny snímaly aktivity na Mariánském náměstí v Praze, jsou vygenerovány trajektorie jednotlivých uživatelů prostoru. Promítnutím jejich přesné polohy do půdorysu náměstí získáváme tzv. heatmapy pohybu osob, tedy zobrazení hustoty trajektorií. Výhodou tohoto přístupu je získání velkého množství informací o pohybu v celé ploše náměstí v dlouhém časovém úseku. Oproti standardnímu zkoumání metodou zúčastněného pozorování nebo pouhým počítáním osob, které projdou přes určený práh, jsou tyto informace nezatížené osobou výzkumníka, jsou přesné a mapují prostor jako celek. Pro komplexnější obrázek o využití prostoru jsme vyvinuli sémantický anotátor – nástroj, pomocí kterého výzkumník přiřazuje na základě kamerových záznamů aktivity osob k jednotlivým trajektoriím. Větší automatizace celého procesu by umožnila provádět složitější úlohy o hledání vztahu mezi daným místem a aktivitami lidí v něm.

Detail

doc. Ing. arch. Jana Zdráhalová, Ph.D., Ing. arch. Lukáš Kurilla, Ph.D.

17-V2A. Tracker Modul trackeru zpracovává data z videostreamů nebo uložených nahrávek, ve kterých jsou detekovány jednotlivé objekty (osoby). Pozice osob nalezené v jednotlivých kamerách jsou následně zkombinovány s využitím statistického přístupu a pravděpodobnostních technik pro nalezení robustního odhadu pozice objektu. Výsledkem jsou pozice objektů projektované do společného mapového plánu sledované oblasti. 17-V2B. Geometry and Semantic Annotators Nástroj pro anotaci uložených videí. Je určen pro přímé použití jako prostředek pro přípravu dat k urbanistické analýze. Nástroj umožňuje načítat uložená natrackovaná data, která lze prohlížet a dále manuálně upravovat nebo doplňovat. Alternativně lze použít pro získání ground truth informace pro další vývoj trackovacích systémů. 17-V2C. Crowd Analytics Další modul softwaru - interagent - představuje analytický nástroj pro detekci, charakterizaci a klasifikaci interakčních vzorců mezi agenty, založený na aplikaci komplexních síťových přístupů a nelineární dynamiky na předzpracovaná data z jiných částí, včetně detekce skupin a určování kauzality. Tato část bude schopna na základě zpracovaných dat poskytnout analytické pohledy na aktuální urbanistickou situaci. Tento úhel pohledu následně umožní příslušným odborníkům učinit přesnější urbanistické rozhodnutí – ICS CAS.

Detail

Ing. arch. Lukáš Kurilla, Ph.D., Ing. arch. Šimon Prokop, Ing. arch. Jiří Vele

15116 Ústav modelového projektování

With the advancement of augmented and virtual reality technologies both in scale as well as accessibility, the Metaverse (Stephenson, 1992, Hughes, 2022) has emerged as a new digital space with potential for the application of architectural creativity and design. With blockchain integration, the concept of the Metaverse shows promise in creating a “decentralised” space for design and creativity with rewards for its participants. As a platform that incorporates these technological components, does the Metaverse have utility for architectural design? Is there something truly novel in what the Metaverse brings to architectural computing, and architectural design? The paper constructs a qualitative knowledge graph that can be used for the evaluation of various kinds of Metaverses in and for architectural design. We use Design Science Research methods to develop the knowledge graph and its evaluative capacity, stemming from our experience with two Metaverses, Decentraland and Cryptovoxels. The paper concludes with a discussion of knowledge and practice gaps that are evident, framing the opportunities that architects might have in the future in terms of developing Metaverse(s).

Detail