As the international platform for the digitalisation of the construction, real estate and facility management industries, BIM World MUNICH is able to draw on a large number of experts to provide insights into current developments in the market.
One of these experts is Stefan Kaufmann, Product Manager at ALLPLAN responsible for BIM Strategy & New Technologies.
We asked him eight questions about the role of artificial intelligence in the construction industry and the opportunities and challenges it presents.
What role do AI-supported BIM models play in optimizing construction processes to reduce CO₂ emissions?
AI-based methods currently play a minor role in planning and construction. This will change in the near future. Statistical methods such as machine learning have the potential to increase efficiency in many areas where logical methods have so far reached their limits. This applies equally to the planning of new buildings and refurbishment projects. Creating a life cycle assessment on the basis of a BIM model is trivial from a mathematical point of view. What is difficult is creating the models and enriching them with characteristic values for sustainability assessment. In future, BIM-based processes will be able to generate BIM models from point clouds or 2D as-built documents. The assignment of components to material classes can be done automatically with AI, as can the assignment of CO2 parameters. This facilitates access to relevant information and increases the competence of planners,
To what extent can AI and digital twins contribute to the circular economy in the construction industry?
In addition to AI-based BIM modeling and life cycle assessment, new methods could emerge that make it possible to better predict the condition and reliability of components. Today, sensors installed with statically relevant components already provide time series of measurement data on the basis of which AI-based systems can determine spontaneous anomalies and long-term degradation of the load-bearing behavior. In case of doubt, non-destructive measurement methods in combination with AI-based evaluation will play an increasingly important role in the future in order to be able to make statements about the inner workings of components. Unfortunately, as has been shown in the past, the current testing methods for reinforced concrete have their limits.
To what extent can machine learning help to monitor the energy efficiency of buildings?
Unfortunately, many buildings consume significantly more energy during operation than was forecast during planning. The so-called “performance gap” can be up to 40%. Due to the low energy costs, this was previously not relevant. With the increasing importance of climate change and rising energy costs, real estate operators are now required to monitor their properties more closely. To this end, innovative real estate companies are already using AI processes that check the plausibility of energy consumption at high frequency so that measures can be taken quickly if necessary.
How can CO2 be saved on the construction site with BIM? What role does AI play there?
While 2D plans always leave gaps in the planning, BIM models force everyone involved to fully plan the building. This is time-consuming, but prevents planning errors and execution deficiencies. The fact that more than 20% of the material is removed from the construction site as waste not only has consequences for the profitability of the construction industry, but also for the climate.
Modern AI processes enable the autonomous tracking of targets. This is particularly advantageous in the field of robotics. Robots act much more intelligently as a result, making laborious training easier or unnecessary. In addition, AI is increasingly able to quickly understand the context in which machines operate. This makes the use of automation solutions in prefabrication and on the construction site cheaper and more reliable.
What role does the standardization of BIM play in achieving the political goals for more sustainability in the construction industry?
Open standards in the field of BIM are the basis for the broad and long-term use of BIM models. Once a model has been created, for example, it can be used as an IFC file in many applications and software solutions in the planning and execution of buildings to increase sustainability. BIM can also be used effectively in the operating phase. At ALLPLAN, for example, we use a BIM model of our offices to optimize the efficiency of the workstations and thus only need 0.7 workstations per employee. Our ecological commitment in our office conversion was recognized by the DGNB with its highest award, the platinum certificate.
What challenges are there when integrating AI into existing BIM systems?
BIM systems such as ALLPLAN have several thousand functions. The aim is for AI to be able to understand the user’s goals and develop strategies to pursue them autonomously using the available CAD functions. If functions are missing, they are programmed independently by the AI. There are various methods for this, which we are working on as part of research projects.
Which standards and regulations must be observed for the use of AI in BIM?
As a European software company, compliance with all standards relating to AI, copyright and the protection of personal data is particularly important to us. The autonomous programming of software code by AI adds new requirements. We ensure that risk-free use is possible within the framework of the strict requirements of the European legal area in the EU and beyond.
How could the role of architects and civil engineers change as a result of AI-supported BIM systems?
In the future, we will all have the opportunity to delegate increasingly complex tasks to assistants. This entails a particular responsibility and need for care, regardless of whether our assistants are human or artificial intelligence.
What developments can be expected in the future for AI and BIM in sustainable construction?
My vision of the future is that sustainable construction, BIM and AI will become the “new normal”. In the future, people will wonder about the careless use of the hard-to-circulate and material-intensive building structures that we are leaving to the next generations. AI assistants will be everywhere, enabling us to act more competently, plan more creatively and produce better buildings more cheaply. BIM models will be generated in a highly automated way and will be available to us for every existing and planned building. Many BIM applications will be replaced by conversational analytics. 2D plans will continue to exist. They will be generated spontaneously from the BIM models depending on requirements, but (hopefully) no longer printed on paper.
Stefan Kaufmann taught and researched at the Technical University of Munich on topics such as BIM, parametrics, digital design and automated production. From 2014 to 2018, he was managing director of the Leonhard Obermeyer Centre, one of the world’s leading research centres for digital methods in construction. As an advisor to the German Federal Ministry of Transport and Digital Infrastructure, he helped develop Germany’s national BIM strategy. Today he is a Product Manager at ALLPLAN, responsible for BIM Strategy & New Technologies, with a particular focus on the integration of artificial intelligence into the design process. He is a member of the advisory board of buildingSMART Germany and part of BIM World MUNICH’s expert committee.