Artificial Intelligence In Construction Sector

As it is evidenced by different volumes within this publication Building & Management, the professionals of the construction sector are adapting in a more generalized way, tools and frameworks that are more collaborative, such as Open BIM, Lean Construction, Last Planner System, Integrated Project Delivery, etc. To the same extent, these new instruments appear in offices and construction sites, thanks to the democratization or extended use of technological development coming from other sectors. The removal of barriers blocking a generalized access to this technology has triggered an increasing of its use.

Technology is influencing almost every industry, and the design and construction fields are starting to adopt all theses advancements. It is no longer odd to fly a drone over a construction site, to visualize a project before completion with virtual reality, or choose the best setting based on predictions made by Artificial Intelligence (AI).

Photo by Max Langelott on Unsplash

The impact of Artificial Intelligence in BIM

Artificial Intelligence (AI) is one of the technologies making great impact due to its future possibilities in the whole construction process. AI can take advantage of the analysis of large sets of data, and in the construction sector, besides the construction element, we also generate a valuable information in the form of data, which is rarely ever used.

AI is associated with a list of terms, such as algorithm, Machine Learning, and automation, which create a fear of the unknown among the professionals. In many fields, not only in the construction industry, there is a true fear towards AI: Will machines become more intelligent than humans? According to Eleni Vasilaki, professor of computation neuroscience, that fear is misplaced. “Ultimately, just because AI can learn, we cannot really conclude that it will learn all of the aspects of human intelligence and become more clever”, said Prof. Vasilaki, adding that we should make sure that we use these technologies in a beneficial way. “The automatic learning and the AI are tools. These can be used in the right way or the wrong way, just like everything else. It is the way in which they can be used that we should be concerned about, not the methods themselves”.

How the application of AI can impact the construction project management, and in particular the BIM project, is still unknown. Designers, architects, and engineers find more questions than answers. What is clear is that the processes for simulation of the building and BIM produce so much data that the majority of the organizations do not know what to do with them.

Hence, it is fundamental to understand the amount of data that is produced in the process of drawing, BIM modelling, construction, and building maintenance. The architects, engineers, and other construction professionals are not using all of this data for their own benefit, or that of their customers. The data stream generated by construction is not usually used, or at least it is not used in the proportion of the possibilities provided by AI.

The tendency in a sector not accustomed to the standardized methods and processes, is to move on to the next project without considering how to use the collected data for improvement. The expert in construction technology Nicholas Klokhol explains the possible implications of AI and Big Data applied to the context of BIM in the construction sector, and its main current problem: once the architectural project is built, 95% of the generated data is either deleted or not properly archived, hampering future analyses and exploitation.

By using data collected from models, simulations, and even through the physical elements as sensors inside finished constructions, we can transform the design process or, at least, innovate with each new construction project. AI provides a new perspective from which we can imagine how to take advantage of the present information, the data, both from previous modelling and from finished buildings.

Machine Learning will automate data analysis processes.

One of the possibilities to deepen the data analyses through AI is associated with the automation of the management processes. The majority of the text referring to the evolution of AI in the direction and management of projects, and to the management of the construction site in particular, derive from a scene in which AI is tasked with the automation and integration of  the information originating in the projects.

According to diverse analyses, like the one by PWC in the article “AI will transform Project management. Are you ready?”, AI will revolutionize every process for data reporting and its subsequent automation through Machine Learning. In this first instance, the applications will combine using the information to different ends, facilitating the data acquisition tasks, or sending alerts when detecting potential problems in the project. In the future, AI might even manage more processes within the projects, from virtual assistants helping the project team improving their work, to machines capable of learning and autonomously manage the projects.

Fuente: PWC. AI will transform project management. Are you ready?

Currently, the estimations made by the project and construction managers are based on the study of historical data on the tasks that have been already performed. This allows the project managers to estimate how long will it take to finish future tasks, as well as the associated costs and risks. For these professionals, registering and analyzing all of this information might become a slow and tricky task; however, by combining an AI-based solution with project management software, we will be able to automate the complete procedure for the analyses of  the historical information. The greater  access to the available information, providing huge quantities of data, the more reliable and compatible with business procedures the AI solution. The AI will save project managers and their organizations valuable time as a result of the automation of repetitive procedures, such as the one mentioned.

For the professionals, this is the first scenario that crosses their imagination: one in which much of the tasks that now occupy their agendas will be replaced by automatic algorithms. The uncertainty for the implicated people is obvious, since it will affect resource assignment, as well as workplaces involved in the planning and management of construction sites.

Further outcomes of the application of AI to BIM could be the improvement of the design itself, as well as the planning of its construction and the demand for resources during the building lifecycle. Several companies are already exploiting the accumulated data from models and constructions for their application in the optimization of conceptual models for future projects. AI techniques use the existing data to predict the optimal functioning of different designs. This way, the designers have the possibility to optimize the development and control the quality of the materials, structures, and installations. Moreover, for the promoters themselves, it assures cost effectiveness of the conceptual models in BIM that might be built in an efficient way and have a life cycle that is cheaper to maintain.

One of the best algorithms for the interpretation and improvement of the building design are the so called “Evolutionary Algorithms” (EA). These are methods for optimization and search of solutions based on postulates of the biological evolution.  EA interprets BIM models through AI mimicking the evolutionary process and provides design solutions that intend to optimize several parameters.

Two of the most reputed experts in AI, Jan Cudzik y Kacper Radziszewski advocate the use of this technology applied to the improvement of the architectonic design. In their article “Artificial Intelligence Aided Architectural Design”, they indicate that EAs are becoming more interesting for artists, designers, and architects. This application is not new and its implications in architecture already have a certain degree of experience. EAs can help drawers solving complex problems like the optimization of different structural typologies or the selection of the best materials for a certain situation. The machine learning and the pattern recognition can analyze big quantities of data in order to obtain useful information. By applying the processing power of the nowadays computers to these massive quantities of data that are available to us, we will continue improving in order to approach more and more to the best possible buildings through the AI.

Other aspects related to the design, and in particular to the energetic performance, illumination, quality of air and hygrothermal comfort, are key aspects that can be improved with the data produced by BIM models and that could considerably improve the operating cost during the building life cycle. By adopting BIM, the building owners can improve the quality of the buildings, highly reduce the costs during the building life cycle, improve the experience of the design projects from the beginning to the end, optimizing the cooperative competition and improving the occupation rates and usage. The obvious usage of BIM models for Facility Managers during the exploitation time of a real estate asset, will be empowered by solutions that are able to maximize the available resources.

The positive impact of Artificial Intelligence on construction projects

Even though we are not able to answer the question of how AI will affect the design and constructive process, we are sure that somehow it will affect and change the way we work and also to which tasks we will dedicate our uptime. Fear of the disappearance of certain roles and technicians will persist, and certainty new specialists will emerge, related to all of these new tools. Nonetheless, it is generally accepted that we will use these new technological capabilities to automate those processes that can be made by a machine, including, maybe, the creative processes. The architect Daniel Dendra, funder of the company Another Architect Studio, an architecture studio that encourages the use of open source designs, believes that the architect might be replaced by a machine using AI in less than 20 years. Daniel does not see this like an unfortunate fact, but as the natural process in which many decisions can be optimized and automated thanks to technology. This fact will provide, at the same time, the opportunity to change the way in which we relate to technology and work, as well as to the reality of a sector in which a lot of professionals will be replaced by algorithms.

We can argue that artificial intelligence has the potential to change the whole discipline, by providing it with new possibilities and ways to explore.

BIM has helped architects, engineers, owners, contractors and construction consultants to improve cooperation and to start speaking in the same language. Besides, BIM has improved both the design processes and the design itself. It has changed the way we work, and the how we dedicate our the time to tasks that bring differential value to the constructive process. With this text being an editorial in which we allow ourselves  certain interpretative liberties from reality and from the future of the profession, we can affirm that the differential value we can contribute has to do with the collaboration that feeds the whole process. And because of that, not only collaboration tools, as BIM or Last Planner System, about which there are some articles in this number, but also the skills of the construction process technicians and managers, will acquire greater importance, if that is possible.

However, we cannot be completely certain about this. As stated at the beginning, the professionals have far more questions than answers, about the AI and their work. Cudzik and Radziszewski state that “we can argue that artificial intelligence has the potential to change the whole discipline, by providing it with new possibilities and ways to explore.” Maybe this will be our new role within the constructive process – to become explorers of new possibilities yet to uncover.

Photo by Franki Chamaki on Unsplash

The present article is part of the Editorial for the number 2019-02 of the scientific magazine  Building & Management.

Comentarios