Generate Floor Plans Research Articles

Floor plan generation: The interplay among data, machine, and designer

Recent advancements in machine learning (ML) in architectural design led to new developments in automated generation of floor plans. However, critical evaluation of ML-based generated floor plans has not progressed proportionally due to the subjectivity and complexity of the assessment, particularly for large and more complex floor plans. Accordingly, a hybrid (quantitative and qualitative) floor plan evaluation scheme is introduced in this study, focusing on multiple architectural aspects. To verify the effectiveness of the proposed framework, the evaluation scheme is applied on the generated floor plans resulting from two baseline computer vision models. The models have been trained on a newly introduced large-scale floor plan dataset called Modified Swiss Dwellings (MSD). The results showed that despite the progression of computer vision models for floor plan generation, they still have difficulty capturing the more complex architectural qualities. In addition, the prospect of floor plan generation and evaluation and possible future developments are discussed.

Automating computational design with generative AI

AbstractAI image generators based on diffusion models have recently garnered attention for their capability to create images from simple text prompts. However, for practical use in civil engineering they need to be able to create specific construction plans for given constraints. This paper investigates the potential of current AI generators in addressing such challenges, specifically for the creation of simple floor plans. We explain how the underlying diffusion‐models work and propose novel refinement approaches to improve semantic encoding and generation quality. In several experiments we show that we can improve validity of generated floor plans from 6% to 90%. Based on these results we derive future research challenges considering building information modeling. With this we provide: (i) evaluation of current generative AIs; (ii) propose improved refinement approaches; (iii) evaluate them on various examples; (iv) derive future directions for diffusion models in civil engineering.

Intelligent floor plan design of modular high-rise residential building based on graph-constrained generative adversarial networks

In the context of Modular High-Rise Residential Buildings (MHRBs), designing floor plans involves intricate complexities due to the need for adherence to numerous domain-specific design rules. To address this issue, our research introduces a novel framework based on a Graph-Constrained Generative Adversarial Network (GC-GAN) specialized for generating MHRB floor plans. This enhanced GC-GAN incorporates knowledge graphs that encapsulate domain-specific constraints and guidelines, thereby generating floor plans that exhibit realism, diversity, and conformity to established design principles. Additionally, the framework integrates a sophisticated image-to-vector conversion algorithm that enables seamless alignment with a predefined flat-design standardization library. A salient feature of this framework is the automated generation of Building Information Modeling (BIM) models, which rigorously conform to the modularity specifications essential for efficient modular construction. The efficacy and practical applicability of our approach have been validated through an exhaustive analysis covering fifteen cases across five diverse scenarios.

Automated layout of modular high-rise residential buildings based on genetic algorithm

The conventional design process of modular high-rise residential buildings (MHRBs) heavily relies on the experience of designers, leading to concerns about design efficiency and quality. Although some methods have been suggested for generating floor plans, there is a scarcity of studies focusing on the automated layout of MHRBs. This paper utilizes the genetic algorithm (GA) to automatically generate MHRB layouts. The approach consists of two stages within the GA framework: the first stage handles the layout design of modular flats, while the second stage is dedicated to the layout design of the traffic core. To facilitate the process, computer-aided design (CAD) drawings are automatically generated using the Autodesk CAD application programming interface, utilizing the information acquired from the proposed approach. Additionally, a website is developed to enable interaction with designers, and two examples are provided to demonstrate the effectiveness of the proposed approach.

An integrated method of automated layout design and optimization for modular construction

PurposeThis paper aims to propose a generative design method combined with meta-heuristic algorithm for automating and optimizing the floor layout of modular buildings using typical standardized module units, which are the room module, the corridor module and the stair module.Design/methodology/approachThe integrated framework involves the generative design method and optimization for modular construction. The generative rules are provided by geometric relationships and functionalities of the module units. An evaluation function of the generated floor plans is also presented by the combination of project cost and cost penalties for the geometric features. The multi-population genetic algorithm (MPGA) method is provided for the optimization of the combination of costs.FindingsThe proposed MPGA method is demonstrated fast and efficient at discovering the globally optimal solution. The results indicate that when the unit price of modules is high, the transportation distance is long, or the land cost is high, the layout cost, which related to the symmetry, the compactness and the energy is tend to be lower, making the optimal layout economical.Originality/valueThis paper presented an integrated framework of generative floor layout and optimization for modular construction by using typical module units. It fulfills the need for automated layout generation with repetitive units and corresponding assessment during the early design stage.

Generating floor plans with deep learning: A cross-validation assessment over different dataset sizes

The advent of deep learning has enabled a series of opportunities; one of them is the ability to tackle subjective factors on the floor plan design and make predictions though spatial semantic maps. Nonetheless, the amount available of data grows exponentially on a daily basis, in this sense, this research seeks to investigate deep generative methods of floor plan design and its relationship between data volume, with training time, quality and diversity in the outputs; in other words, what is the amount of data required to rapidly train models that return optimal results. In our research, we used a variation of the Conditional Generative Adversarial Network algorithm, that is, Pix2pix, and a dataset of approximately 80 thousand images to train 10 models and evaluate their performance through a series of computational metrics. The results show that the potential of this data-driven method depends not only on the diversity of the training set but also on the linearity of the distribution; therefore, high-dimensional datasets did not achieve good results. It is also concluded that models trained on small sets of data (800 images) may return excellent results if given the correct training instructions (Hyperparameters), but the best baseline to this generative task is in the mid-term, using around 20 to 30 thousand images with a linear distribution. Finally, it is presented standard guidelines for dataset design, and the impact of data curation along the entire process.

A generative design method of building layout generated by path

Abstract The floor plan is a key part of architectural design and has the characteristics of multi-objective evaluation. Traditional methods are often laborious and may lead to re-work due to optimisation, which affects efficiency. Related research uses various algorithms to generate floor plans to improve efficiency. Based on the leading role of architects on floor plans, this paper proposes a generative design method that combines the design process. It takes the space shape as the starting point, simplifies the design process into a mathematical model, and uses Rhino and Grasshopper to complete the algorithm development. Using different algorithm combinations, a large number of new layouts can be explored, and building layouts can be adjusted freely through data optimisation without having to redesign. The design evolution of floor plans of different cases shows that such a generative design method is feasible, reusable, and more efficient than traditional methods.

Walk2Map: Extracting Floor Plans from Indoor Walk Trajectories

AbstractRecent years have seen a proliferation of new digital products for the efficient management of indoor spaces, with important applications like emergency management, virtual property showcasing and interior design. While highly innovative and effective, these products rely on accurate 3D models of the environments considered, including information on both architectural and non‐permanent elements. These models must be created from measured data such as RGB‐D images or 3D point clouds, whose capture and consolidation involves lengthy data workflows. This strongly limits the rate at which 3D models can be produced, preventing the adoption of many digital services for indoor space management.We provide a radical alternative to such data‐intensive procedures by presenting Walk2Map, a data‐driven approach to generate floor plans only from trajectories of a person walking inside the rooms. Thanks to recent advances in data‐driven inertial odometry, such minimalistic input data can be acquired from the IMU readings of consumer‐level smartphones, which allows for an effortless and scalable mapping of real‐world indoor spaces. Our work is based on learning the latent relation between an indoor walk trajectory and the information represented in a floor plan: interior space footprint, portals, and furniture. We distinguish between recovering area‐related (interior footprint, furniture) and wall‐related (doors) information and use two different neural architectures for the two tasks: an image‐based Encoder‐Decoder and a Graph Convolutional Network, respectively. We train our networks using scanned 3D indoor models and apply them in a cascaded fashion on an indoor walk trajectory at inference time.We perform a qualitative and quantitative evaluation using both trajectories simulated from scanned models of interiors and measured, real‐world trajectories, and compare against a baseline method for image‐to‐image translation. The experiments confirm that our technique is viable and allows recovering reliable floor plans from minimal walk trajectory data.

Computer-aided approach to public buildings floor plan generation. Magnetizing Floor Plan Generator

For architects as well as developers and urban planners working on the floor plans or estimating the shape and dimensions of large buildings is always a challenge. This task requires some knowledge and what is more important – even with understanding of the process it is a very time-consuming task. One should take into consideration the arrangement of all rooms as well as adjacencies and connections of main spaces. Presented project can be considered as an exploration of various ways of generating floor plans for public buildings, which was followed by creating a new algorithm for solving that task. Public buildings were chosen as a main target of conducted research because of their complex and non-standardized structure. The aim was to try different previously described approaches, invent completely new techniques and methods, choose the best of them and incorporate them into our own generator. The creation of working generator prototype was an important target of the project and should not be considered as its final stage, since this trial revealed some positive and negative sides of chosen approach and therefore serves as an intermediate stage. Floor plans analysis was conducted in order to understand the structural patterns of communication spaces in built public houses. One of positive outcomes of this investigation is that public buildings do not share a lot of features in common, therefore this field of exploration is rather unobstructed and provides a lot of possibilities for experiments.

Data-driven interior plan generation for residential buildings

We propose a novel data-driven technique for automatically and efficiently generating floor plans for residential buildings with given boundaries. Central to this method is a two-stage approach that imitates the human design process by locating rooms first and then walls while adapting to the input building boundary. Based on observations of the presence of the living room in almost all floor plans, our designed learning network begins with positioning a living room and continues by iteratively generating other rooms. Then, walls are first determined by an encoder-decoder network, and then they are refined to vector representations using dedicated rules. To effectively train our networks, we construct RPLAN - a manually collected large-scale densely annotated dataset of floor plans from real residential buildings. Intensive experiments, including formative user studies and comparisons, are conducted to illustrate the feasibility and efficacy of our proposed approach. By comparing the plausibility of different floor plans, we have observed that our method substantially outperforms existing methods, and in many cases our floor plans are comparable to human-created ones.

Customization and generation of floor plans based on graph transformations

Abstract This paper introduces an approach for automatic generation of rectangular floor plans based on existing legacy floor plans with the capability of further improvement and customization. Our approach first derives a dual graph from the given input file specifying a floor plan. It then either automatically reproduces varied floor plans retaining the connectivity of the original plan or performs transformation rules to manipulate spatial relations among rooms, and to generate modified floor plans corresponding to specific requirements. Our approach introduces constraints, such as the maximum width-height ratio, to support the flexibility for various design requirements. A graphical user interface is provided for users to perform the automatic generation process. An experiment has been conducted to validate the feasibility of our approach and time taken in generating floor plans. It shows that our method is able to generate highly-customized floor plans in reasonable time.

THE GENERATION OF BUILDING FLOOR PLANS USING PORTABLE AND UNMANNED AERIAL VEHICLE MAPPING SYSTEMS

Indoor navigation or positioning systems have been widely developed for Location-Based Services (LBS) applications and they come along with a keen demand of indoor floor plans for displaying results even improving the positioning performance. Generally, the floor plans produced by robot mapping focus on perceiving the environment to avoid obstacles and using the feature landmarks to update the robot position in the relative coordinate frame. These maps are not accurate enough to incorporate to the indoor positioning system. This study aims at developing Indoor Mobile Mapping System (Indoor MMS) and concentrates on generating the highly accurate floor plans based on the robot mapping technique using the portable, robot and Unmanned Aerial Vehicles (UAV) platform. The proposed portable mapping system prototype can be used in the chest package and the handheld approach. In order to evaluate and correct the generated floor plans from robot mapping techniques, this study builds the testing and calibration field using the outdoor control survey method implemented in the indoor environments. Based on control points and check points from control survey, this study presents the map rectification method that uses the affine transformation to solve the scale and deformation problems and also transfer the local coordinate system into world standard coordinate system. The preliminary results illustrate that the final version of the building floor plan reach 1 meter absolute positioning accuracy using the proposed mapping systems that combines with the novel map rectification approach proposed. These maps are well geo-referenced with world coordinate system thus it can be applied for future seamless navigation applications including indoor and outdoor scenarios.

THE GENERATION OF BUILDING FLOOR PLANS USING PORTABLE AND UNMANNED AERIAL VEHICLE MAPPING SYSTEMS

Abstract. Indoor navigation or positioning systems have been widely developed for Location-Based Services (LBS) applications and they come along with a keen demand of indoor floor plans for displaying results even improving the positioning performance. Generally, the floor plans produced by robot mapping focus on perceiving the environment to avoid obstacles and using the feature landmarks to update the robot position in the relative coordinate frame. These maps are not accurate enough to incorporate to the indoor positioning system. This study aims at developing Indoor Mobile Mapping System (Indoor MMS) and concentrates on generating the highly accurate floor plans based on the robot mapping technique using the portable, robot and Unmanned Aerial Vehicles (UAV) platform. The proposed portable mapping system prototype can be used in the chest package and the handheld approach. In order to evaluate and correct the generated floor plans from robot mapping techniques, this study builds the testing and calibration field using the outdoor control survey method implemented in the indoor environments. Based on control points and check points from control survey, this study presents the map rectification method that uses the affine transformation to solve the scale and deformation problems and also transfer the local coordinate system into world standard coordinate system. The preliminary results illustrate that the final version of the building floor plan reach 1 meter absolute positioning accuracy using the proposed mapping systems that combines with the novel map rectification approach proposed. These maps are well geo-referenced with world coordinate system thus it can be applied for future seamless navigation applications including indoor and outdoor scenarios.

TOWARDS OBJECT DRIVEN FLOOR PLAN EXTRACTION FROM LASER POINT CLOUD

During the last years, the demand for indoor models has increased for various purposes. As a provisional step to proceed towards higher dimensional indoor models, powerful and flexible floor plans can be utilised. Therefore, several methods have been proposed that provide automatically generated floor plans from laser point clouds. The prevailing methodology seeks to attain semantic enhancement of a model (e.g. the identification and labelling of its components) built upon already reconstructed (a priori) geometry. In contrast, this paper demonstrates preliminary research on the possibility to directly incorporate semantic knowledge, which is itself derived from the raw data during the extraction, into the geometric modelling process. In this regard, we propose a new method to automatically extract floor plans from raw point clouds. It is based on a hierarchical space partitioning of the data, integrated with primitive selection actuated by object detection. First, planar primitives corresponding to vertical architectural structures are extracted using M-estimator SAmple and Consensus (MSAC). The set of the resulting line segments are refined by a selection process through a novel door detection algorithm, considering optimization of prior information and fitness to the data. The selected lines are used as hyperlines to partition the space into enclosed areas. Finally, a floor plan is extracted from these partitions by Minimum Description Length (MDL) hypothesis ranking. The algorithm is applied on a real mobile laser scanner dataset and the results are evaluated both in terms of door detection and consecutive floor plan extraction.

TOWARDS OBJECT DRIVEN FLOOR PLAN EXTRACTION FROM LASER POINT CLOUD

Abstract. During the last years, the demand for indoor models has increased for various purposes. As a provisional step to proceed towards higher dimensional indoor models, powerful and flexible floor plans can be utilised. Therefore, several methods have been proposed that provide automatically generated floor plans from laser point clouds. The prevailing methodology seeks to attain semantic enhancement of a model (e.g. the identification and labelling of its components) built upon already reconstructed (a priori) geometry. In contrast, this paper demonstrates preliminary research on the possibility to directly incorporate semantic knowledge, which is itself derived from the raw data during the extraction, into the geometric modelling process. In this regard, we propose a new method to automatically extract floor plans from raw point clouds. It is based on a hierarchical space partitioning of the data, integrated with primitive selection actuated by object detection. First, planar primitives corresponding to vertical architectural structures are extracted using M-estimator SAmple and Consensus (MSAC). The set of the resulting line segments are refined by a selection process through a novel door detection algorithm, considering optimization of prior information and fitness to the data. The selected lines are used as hyperlines to partition the space into enclosed areas. Finally, a floor plan is extracted from these partitions by Minimum Description Length (MDL) hypothesis ranking. The algorithm is applied on a real mobile laser scanner dataset and the results are evaluated both in terms of door detection and consecutive floor plan extraction.

Effective mobile mapping of multi-room indoor structures

We present a system to easily capture building interiors and automatically generate floor plans scaled to their metric dimensions. The proposed approach is able to manage scenes not necessarily limited to the Manhattan World assumption, exploiting the redundancy of the instruments commonly available on commodity smartphones, such as accelerometer, magnetometer and camera. Without specialized training or equipment, our system can produce a 2D floor plan and a representative 3D model of the scene accurate enough to be used for simulations and interactive applications.

計画与条件に適合する長方形分割図の作成方法について : グラフ理論的アプローチによる平面計画の方法 その1

The purpose of this study is to develop a logical method generating floor plans by applying Graph Theory. An outline of the method proposed in this paper is as follows. 1) The problem of generating floor plans is formulated as a bunch of sub-problems, such as a problem of transformation from a non-planar graph to a planar one, a problem of embedding inner vertices into faces of a graph, a problem of assigning feasible orientations to each room in a circular graph. 2) The method is useful to find out all the plane graphs satisfying given adjacency and out-facing requirements by introducing outer Hamilton graph. 3) It is also useful to generate all the floor layout graphs and rectangular dissections efficiently by a searching technique with constraints, that is, feasibility of selecting orientations and of being a rectangular dissection.