In early September, Korea's leading construction companies will come together to discuss the future of the construction industry through the application of digital technologies. The companies are expected to present real-world examples of their smart construction technologies, including digital twins, BIM (building information modeling), and AI(artificial intelligence). As AI has begun to impact many industries, the construction industry has been no exception. Recently, LLMs such as ChatGPT have also been actively introduced. In this article, we'll take a look at why AI in construction is important, real-world examples, and the technologies used, as well as their advantages and disadvantages.
Why architectural AI is important
Design and automate architecture in new ways
The mathematical and physical demands of architectural engineering used to be the result of pure, creative imagination. Architects would consider the pros and cons of various options to decide which materials to use for a particular partof a building. Even today, many architects rely on experience and manual analysis to make these decisions. This has led to architecture being seen as a human endeavor, but things are different now.
By using AI in architecture, youcan automate parts of the same design process. Because AI specializes in understanding large amounts of data, software can be used to compare numerous materials in a very short amount of time. Based on this analysis, designers candraw conclusions about their quality and suitability for a particular space.Automated architectural design also allows architects to take a more quantitative and objective approach.
And thanks to generative AI like computer-aided design(CAD), DALL-E-2, Midjourney, and Stable Diffusion, new designs can be improved and informed. Using AI, architects can get design ideas for a wider variety of shapes.
Improve safety and productivity on construction sites
Safety on construction sites is atop priority. AI technology can help you better monitor and manage the safety of your workers. For example, you can use object detection technology to monitor the wearing of personal protective equipment in real time and detect hazards. You can also monitor hazardous areas and identify hazards to provide warnings and instructions to workers to help prevent accidents.
Image processing and pattern recognition algorithms can also be utilized to monitor the health of structures and identify defects (cracks, corrosion, etc.). This information can be utilized for regular maintenance planning and early preventive measures, helping to extend the life of the structure and improve its reliability.
AI offers many ways to significantly improve productivity in the construction process. AI systems connected to sensor networks can collect and analyze data in real time to determine current conditions, predict problems, and support decision-making.Automated scheduling and resource management systems can optimize work schedules and resource allocation to improve project progress and efficiency.Additionally, AI-powered architectural design and simulation tools can proactively detect and correct design errors, reducing rework costs and time
ArchitecturalAI Artificial Intelligence Use Cases
A NewArchitectural Design Paradigm (1) - Autodesk
Autodesk's generative designtechnology, Dreamcatcher, leverages AI to generate optimized design solutions:you set the design goals and constraints, and the AI algorithm explores all possible designoptions and presents the best solution.
You set the goals and constraintsfor your design. These can be specific material usage, structural strength,manufacturing methods, cost, etc. The AI algorithm explores all possible designvariations and generates different solution candidates based on the setrequirements and constraints.
The generated design options areevaluated based on performance metrics, and the user can choose the best designsolution from among them. Finally, the selected solution is tested in areal-world environment and its performance is validated. Further modificationsor optimizations can be made as needed, and the process is repeated.
Autodesk's Dreamcatcherautomatically generates optimal design solutions within the goals andconstraints set by the designer, which can save significant time and effortduring the architectural design process.
What is GAN, the hottest generation model right now?
A New Architectural Design Paradigm (2) - Plug city
Fluxity is a leader in digital twin technology, which precisely recreates real-world spaces as three-dimensional virtual environments. The company digitally transforms buildings, basements, subways, airports, factories, resorts, plants, cities, and many other real-world environments to enable contactless remote operations.
Fluxity's digital twin technology is being used extensively on construction sites. Everything from access control to real-time process status and heavy equipment movement routes can be realized in 3D for more systematic management. This is exemplified by Fluxity's work with Hovan Construction to control actual construction sites such as Seoul's Gaebang 5 District and Wirye 9 Block.
Fluxity is focused on seamlessly connecting the physical and digital worlds through digital-twin technology, providing a more efficient and secure operating environment. The company's innovative technology is making a significant impact and proving its value in a variety of industries.
ImprovingConstruction Site Productivity (1) - Hyundai Engineering
Hyundai Engineering has completed the development of the industry's first automatic design system for plant steelstructures using AI machine learning, and has applied for a patent for it. The system can significantly reduce the time required for the design process, contributing to reducing the workload of engineers and improving work efficiency.
The ability to suggest the optimal shape of a structure using AI machine learning algorithms has already made it possible to review constructability at the design stage. This minimizes errors between design and construction, leading to shorter construction periods and a reduction in design costs of about 20% or more. Furthermore, we have developed an image recognition system that automatically recognizes and data hibernating images with AI, and are applying it in practice. In addition, we are utilizing AI technology in various ways, such as prediction based on project data learning, to infer the volume of projects to be bid.
Hyundai Engineering is actively pursuing a variety of AI challenges, including 3D automation, optimized design, and digital twins, with a long-term roadmap to 2025. This continuous R&D and growth ambition is helping to expand the use of AI technologies such as deep learning and machine learning with in the construction industry. These efforts will play a crucial role in improving productivity, reducing costs, and enhancing safety across the construction industry.
IncreasingConstruction Site Productivity (2) - Dusty Robotics Construction AutomationSystems
Dusty Robotics is a cutting-edge robotic solution for streamlining the construction process. It enables a process that is approximately 60% faster than traditional building methods and has achieved 95% accuracy in the placement of building elements. This makes it an excellent choice for anyone in the construction industry who wants to maximize efficiency, reduce human error, and save time on their projects. It is especially suitable for residential and commercial projects.
The Dusty robot is the embodiment of advanced engineering and cutting-edge technology, utilizing state-of-the-art sensors, cameras, and algorithms to navigate complex construction environments with ease. Equipped with light detection and ranging (LiDAR) technology, robots can create detailed maps of construction sites to understand their surroundings and intelligently avoid obstacles.
This level of autonomy reduces the need for human intervention in layout tasks and greatly minimizes the possibility of errors, ultimately leading to more accurate and precise construction results. The bottom line is that it's a much faster process than traditional methods, highly accurate, applicable to a wide range of project types, and reduces long-term costs.
AI technology applied to architectural design
AI is revolutionizing the field of architectural design, providing benefits such as increased productivity, reduced costs, and improved efficiency. In this article, we'll explore AI technologies applied to architectural design.
- Machine learning, Deep learning: Machine learning is a branch of AI, the science of developing algorithms and statistical models that allow computers to learn without being explicitly programmed. Deep learning is a branch of machine learning that uses artificial neural networks to learn complex patterns. These techniques are used to solve optimization problems in the architectural design phase, prevent safety incidents in the construction phase, and more.
- Natural Language Process (NLP): NLP is a technology that allows computers to understand and produce human language. In architecture, it's used for document automation, contract review and management, and more.
- Reinforcement learning: Reinforcement learning is an AI methodology in which a system learns to behave in a way that maximizes its reward in a given environment. It is used for complex simulation problems or robotics applications (e.g., automated construction robots).
- Generative Design: Generative Design is a methodology in which AI explores all possible design solutions within a given set of conditions and suggests the optimal outcome. It has applications in optimizing structural design and improving energy efficiency, among others.
AI technology in construction
In addition to the technologies mentioned above, computer vision (CV), an AI technology that understands image or video data, is playing a key role in the construction sector. It is being used for site monitoring, hazard detection, and automated inspection and maintenance through drones and CCTV. In particular, object detection is helping to ensure industrial safety and productivity by specifically identifying things that managers cannot see in the field.
What is Segment Anything (SAM), a model that allows you to categorize any object with a single click?
Personal protective equipment detection
With deep learning, you can reliably detect a variety of personal protective equipment, including hardhats, safety glasses, gloves, high-visibility jackets, earplugs, and more. This helps ensure the safety of your workers.
Detecting workplace hazards
Technologies like object tracking can be used to identify hazards on construction sites, such as flammable materials, or to monitor workers in hazardous areas. AI models can provide real-time alerts to managers and workers to prevent accidents.
For example, IRIS's AI creates virtual maps and analyzes the entire work area using CCTV camera feeds.Whenever a workplace hazard poses a threat to your workers, you can identify the risk and configure real-time notifications and alerts on WhatsApp, SMS, buzzer, or PA system.
Everguard is also complementing computer vision with wearable technology and sensor fusion to improve worker safety with PPE detection, fall detection, forklift safety, collision avoidance, crane detection, and posture detection.
However, some workers may perceive this as "surveillance" in the name of risk detection. With that in mind, IntensiveEye's visual data is deleted after processing, so no personal information is stored, and their algorithm uses body features to identify human "objects" instead of facial recognition
Corrosion Detection
AI uses algorithms to identify different types of corrosion. It determines whether there is a loss of coating or direct corrosion on a metal, for example, and categorizes it according to severity. For example, if you can distinguish between coating loss, where the protective coating on a metal like steel starts to break down, and actual rust, your workers will be able to react more quickly.
Infosys trained the algorithm with a 70% test train split on the labeled dataset to create 3,000 annotations. It was then able to distinguish between coating loss severity levels (P1, P2, P3, and normal) using four annotation classes.
Architecture with Transformer Structures, What is ViT?
Inspecting infrastructure assets
Remote monitoring allows construction sites to track data about infrastructure assets in real time and manage maintenance schedules. This gives construction managers the ability to track the progress of the overall project and the information and location of individual assets.
Pipeline Inspection
After vehicles (ROVs) and drones capture pipeline survey video, AI models can leverage image processing and optical character recognition (OCR) technology to help accurately detect and classify damaged areas and boundaries.
Fault detection and infrastructure inspection
AI algorithms are utilized to detect defects (cracks, rust, etc.) from image data. These systems are useful for solving infrastructure-related problems, such as assessing road quality or monitoring structures.
Trained on 1,050 images and tested on 450 images, the 'AlexNet' CNN algorithm was able to detect cracks in noisy images, including cracks with shadows and stains, and cracks on rusty and rough surfaces. It achieved a whopping 98.4% accuracy rate.
A company called Eye ForInfrastructure was also able to use AI models to analyze the quality and deterioration of roads. These kinds of systems can be used to monitor infrastructure for problems such as damaged structures, tree overgrowth, drainage issues, and downed power lines.
Advantages and limitations of architectural AI
Artificial intelligence is rapidly advancing and showing great promise in many fields, but the architecture industry is particularly excited about its evolution. This is because the process of architectural design and implementation, which has traditionally been time-consuming and labor-intensive, is being transformed by the introduction of AI. However, there are obvious advantages and limitations.
Advantages
- Increased efficiency: AI can save time and money at many stages of the building process, including design, planning, and implementation. For example, AI can use advanced analytics to quickly find optimal design solutions, or automated construction robots to reduce labor hours.
- Increased predictive power: AI is utilized to learn complex data patterns to predict future trends or issues. This is used to predict maintenance needs, detect safety incident risks, and more.
- Design innovation: AI techniques like Generative Design explore all possible design solutions within a given set of conditions and suggest the optimal outcome. This can lead to innovative design ideas that you may not have thought of before.
- Enable smart buildings and smart cities: The combination of AI and IoT technologies optimizes building and city operations, improves energy efficiency, and makes life easier.
Limits
- Data dependency: AI algorithms rely on large amounts of accurate data, but sufficient and accurate data is not always available for all situations and conditions.
- Cost: Adopting and maintaining a high-quality AI solution can be a significant expense, which can be prohibitive for small to mid-sized architecture firms.
- Regulatory and legal issues: Issues such as privacy are leading to increased regulation around data use. This places constraints on the collection and utilization of data by AI.
- Lack of human element: AI can't fully understand or reflect human creativity, intuition, and complex emotions, so in fields where the human element is critical, such as architectural design, AI can't do it all.
The bottom line: AI in the construction industry can improve construction productivity, reliability, and collaboration.
In the building industry, AI is being leveraged throughout the lifecycle of a building, from the design phase to construction and maintenance. In particular, generative design, building information modeling (BIM), automation and robotics, virtual and augmented reality, and more are being applied to improve productivity and reduce costs.
AI plays a huge role in solving complex problems through big data analysis, which can be used to suggest optimized design solutions, manage processes, monitor safety, and more. These processes used to be done manually by humans, but now AI can do them more accurately and quickly. As a result, architects' proficiency with AI-enabled apps has become highly sought after in recent years, and the architectural design industry in the U.S. and Europe has said that architects with software skills are highly sought after.
However, the adoption and utilization of AI technology, which is still in its infancy, still faces many challenges. For example, the active use of AI in the building industry has raised concerns about the absence of the human element, such as human creativity and intuition. In addition, there are legal compliance and ethical issues to consider, as well as ensuring accuracy and reliability. Nevertheless, the potential for advancing AI technology within the building industry is enormous.Therefore, when introducing artificial intelligence (AI) into the building industry, research and development on technical advancements, data protection and management, and harmonization with human factors will need to be continued, taking into account both the benefits and limitations.
Reference.
- Artificial Intelligence and Architecture
- AI architecture in a minute, an architect's review?
- Using AI to Optimize Construction Design
- Design an apartment in 30 minutes... AI, drones, and robots are changing the construction game | The Korea Economic Daily
- Dusty Robotics - A Fascinating Comprehensive Guide - DotCom Magazine-Influencers And Entrepreneurs Making News
- How can AI be used in construction companies?
- "AI improves construction efficiency" Hyundai Engineering promotes conversational AI development - IT News > Company > Proptech
- 7 Job-ready AI Applications in Construction
.svg)
