In the early 2000s, the Architecture, Engineering, and Construction (AEC) industry had different stakeholders working in separated spaces, for different timeframes across the project lifecycle and not necessarily in a synchronized fashion. This disconnected, linear format is the traditional sequential engineering process, which compromises the inter-team cohesion and project time. With the adoption of the Building Information Modelling (BIM), concurrent engineering superseded sequential engineering, and it has successfully sorted out many pain points of the AEC industry which includes risk mitigation, process optimization, quality improvement, better communication, and visualization.
We’re on the brink of the Artificial Intelligence (AI) revolution across all the industries and the AEC industry is no exception. Even though the adoption of BIM has successfully eliminated many challenges of the traditional methods, the former is not without its own set of limitations. That’s where the need and scope for the adoption of AI in BIM arises. The timely adoption of AI and Machine Learning (ML) can bring about remarkable improvements to existing workflows of the AEC industry by facilitating optimized designs, efficient performance, and quicker project delivery.
With the advent of IoT-enabled smart buildings and digital twin technologies in BIM, there is no shortage of data or information required to stimulate, fuel, and power AI and ML in the industry. The BIM process provides an ample amount of data which is mainly classified into project data and construction data. The project data can be easily acquired with project handbooks, contracts, planning, and Enterprise Resource Planning (ERP) tools. Acquiring construction data, although not as easy as project data acquisition, is possible with the adoption of technologies like IoT and smart buildings. Sensors integrated with smart buildings produce the most critical data required for the machine learning processes and influence decision-making throughout the product and project life cycle. The digital twin adoption alongside IoT in smart buildings increases the production of data, which enhances the scope of AI and ML.
Through Machine Learning, the produced data can be used to improve the efficiency of the process. For instance, In the design phase, by feeding parameters like room loads, supply and return capacity of the room and AHU (Air Handling Unit), number, and the placement of air terminals, the duct routings can be generated automatically. In the operation phase, the neural networks can be trained with data of the HVAC power used across the building asset with related parameters like thermal conductivity and thickness of different ducts and calculate the required heat resistance, the optimal temperature required for the room by learning real-time Internal and external loads and other parameters required.
What if, a smart building is 3D printed with virtual design deliverables of the project and the construction process is carried on with robots with the help of AI and ML. This may sound fantastical, but with the adoption of IoT, Digital twin, AI, ML, and 3D printing disrupting the dynamics of the AEC industry, such breakthroughs are possible. The future of the industry is about to change for good and in hindsight, it will be good for those who are willing to learn and develop themselves.
The National Digital Twin programme (NDTp) was launched in the UK to set up digital twins for all the cities in the UK in order to improve management, operations, and eventually, the overall infrastructural performance and capability. These modern, smart communication channels between the cities, buildings, and an infrastructural approach mark the dawn of a new revolution in the AEC industry.
