Opinion of the Student on Artificial Intelligence in the Built Environment

Editie: 29 - Artificial Intelligence in the Built Environment

Published on: 28 maart 2022

---

 

Artificial Intelligence and Building Physics – Anne Offermans

Within the discipline of Building Physics and Services, there are many opportunities for the implementation of Artificial Intelligence. These opportunities lie within the design phase of a building, as well as during building operations.

When designing a building, the amount of building designs which are generated and simulated are limited to the amount of labor hours available. However, with the implementation of AI, many building designs can be simulated and analyzed within a short period of time. The AI can introduce suggestions to the building designer which might not have been evaluated by the designer him/herself.

Artificial Intelligence can only become as smart as the designer who makes the system.

During the operation phase of a building, AI can help to improve e.g. the indoor air quality, the energy efficiency, and the lighting quality of a building. Smart HVAC systems can be placed within the building, to reduce and increase the heating and ventilation capacity when people are present, or when the indoor air quality is too low. Next to that, AI can be used to check the performance of the installation itself, to check whether maintenance is required to the system. As a result, repairs can be performed on the moment they are required, increasing the energy efficiency of the systems.

However, AI does not work properly without the right implementation. AI can only become as smart as the designer who makes the system. Therefore, a lot of knowledge will be required on both the AI aspect as well as the building physics aspect to make sure that the AI is implemented in the right way.

 

Artificial Intelligence in Urban Design – Sarah Selanno

The emergence of artificial intelligence will inevitably lead to drastic changes within our everyday society. Since the designing of cities is inseparable with the daily urban systems of society, it is without a doubt that in the future AI will change the way we design a city.

AI could be a helpful addition to the designer’s toolkit to predict the behaviour of people by tracking their everyday routing, needs and habits. Uncovering the patterns in society and quantifying them is immensely valuable in urban design.

With the help of AI, the design of cities and places could be optimized. Human errors in misinterpreting the needs of society will be minimalised. Smart city planning could be used to advance smart energy usage and optimise the design of current traffic arteries. Moreover, the usage of AI can lead to better motivated design decision making. All in all, the process of designing can be sped up with the help of immense data sets that are available at any time, tracking the most current trends within the city.

Big Brother is already watching our every step, but we must be careful to not completely surrender to the complacency of AI.

Already, there is a lot of human data available to the public. Whilst the data can be extremely valuable for urban designers, it is frightening when one realises what information is out in the public. As technology evolves and AI progresses, we should take careful notice of the power and the abilities of smart data. Big Brother is already watching our every step, but we must be careful to not completely surrender to the complacency of AI.

AI is a promising innovation and a valuable tool for urban designers, but should be handled with immense precaution and care by authorities, municipalities and design offices.

 

Artificial Intelligence in Structural Design – Olaf Vens

Within the Built Environment, a common misconception about AI is that it equals optimization. This can be seen in a lot of FEM programs, these have solvers which can be used to find an optimum solution. Where this is seemingly close to AI, these solvers are hard coded and can only be adapted by the parameters set by the user.

The power of AI lies in pattern recognition, and problem-solving without a hard set target or path.

The power of AI lies in pattern recognition, and problem-solving without a hard set target or path. A common field within structural design in which AI is used is in the prediction of material properties. With more difficult types of materials it is not the case that you can easily predict their properties. For example fiber reinforced concrete is such a material which is highly dependent on a lot of different parameters and circumstances to determine the properties, an AI can in this case be trained to aid in the prediction of the properties. Still, it is important to note that one should remain aware of what you are doing. It is nowhere near necessary to fully understand the exact underlying mechanics of an AI, but you should be able to ask yourself questions like: how do I process the results or how do I interpret these results. At the moment, for most structural designers in practice, AI will remain a black box and not something we will use a lot. However, with the increasing knowledge, generated amongst others at the TU/e by Hèrm Hofmeyer, the toolbox for AI-assisted design will increase rapidly with more parameters and a larger useable scope, and therefore make it more and more useful. Therefore to conclude, a toolbox to assist in a structural design using AI will probably soon be a useful tool, but for now it is still a work in progress.

 

Artificial Intelligence in Architecture – Max Slangen

In 1973, architectural theorist Charles Jencks created a chart of the architectural movements and trends of the twentieth century. Capturing any history of architecture in one diagram, one glance, is of course a futile task, but Jencks’ diagram is tremendously amusing if nothing more. The section beyond 1973 is especially diverting: Jencks “predicts” an architectural future with laser-pipes, exoskeletons, androids, and design by computer-light pens.

Rationality goes a long way in design, but could never by itself lead to a truly great result.

Despite my efforts, I am yet to find a building with laser-pipes, but it is undeniable that technology has had a deep and lasting impact on the architectural profession. Computers and software are omnipresent and unmissable in every stage of the design and construction process. It has led some designers to speculate about the redundancy of the architect completely in the foreseeable future. Now, if Jencks’ diagram teaches us anything, it is surely that predicting the future is an arduous task, especially if we consider the semi-serious definition of AI reading ‘that which the current computers cannot yet do’. Nevertheless, it seems highly unlikely that the design process could be completely taken over by AI. Design requires creativity and human interaction, with, moreover, a low rate of repetition. Rationality goes a long way in design, but could never by itself lead to a truly great result. A computer may get a perception of space in a purely Cartesian sense, but that is never enough, never satisfactory as the sole interpretation of the world in which we live (and which we design!), as Martin Heidegger so rightfully pointed out.

What AI can do for architects lies much more in the mundane tasks with which we are burdened. BIM has already started this process, but there are worlds to be won when it comes to putting AI to good use in helping us draw, calculate, and automate other repetitive and time-consuming tasks. There is an initial barrier in implementing AI because it tends to have a steep learning curve, but having overcome that, architects could potentially benefit tremendously and spend more time actually designing and considering all the complex decisions in the architectural process AI cannot do for us.

 

Artificial Intelligence in Construction Management – Britt van de Laar

Within the discipline of CME artificial intelligence can be widely used, from an urban to a building scale. To know how artificial intelligence can help the CME track, it is important to know what opportunities AI can bring. Artificial intelligence is a term used for machines and systems to imitate human functions. Within the construction industry buildings, cities and infrastructure can be seen as such systems.

Cities are complex systems with lots of interaction between people, communities and different mobility types. Therefore, on an urban level, mobility and infrastructure are important aspects of CME. When looking at mobility and AI an interesting opportunity lies in the development of self-driving and self-learning cars. Those cars can help to reduce the amount of distraction a driver has during a trip. Those self-learning cars can improve over time their own driving skills. This can, later on, improve the safety on roads and thus in cities. A start has already been made in researching the possibilities of this. So, when this technique will be implemented more often, the learning curves of these cars will also increase. Related to this, city planning will also change when more data about cars and traffic flow will become available. Making it also interesting for other disciplines, like Urban Systems & Real Estate and Urbanism.

Artificial Intelligence can be the impulse the construction industry needs to more innovation.

But not only is artificial intelligence interesting for the mobility part of CME, but also for the project management side. A great example of this is a digital twin that is able to predict. As a project manager, it can be seen beforehand what kind of influences your decisions have on a project. This can be on a building level, modelling from the start of construction, making the foundation and all the phases following that. But also on an urban scale, plan how the construction process will influence the surrounded city area. Also, the time plan of a project can be changed beforehand if delays are already predicted, improving the construction process.

For both examples, AI will accelerate innovations in the construction industry. Artificial Intelligence can be the impulse the construction industry needs to more innovation.

 

Mail the editors