The Commanding Presence of Artificial Intelligence and Digital Twin Technology in AEC

30 April 2024
The Commanding Presence of Artificial Intelligence and Digital Twin Technology in AEC

Minds over machines

Seventy-three years ago, Alan Turing assessed a machine’s capacity to demonstrate intelligent behavior, a notion reflective of modern artificial intelligence. Jumping ahead to 2023, Modern AI systems are revolutionizing the way we approach tasks, make decisions, and interact with technology, ushering in an era where the boundaries between human capabilities and artificial intelligence continue to blur.

What is Artificial Intelligence?

AI is defined as the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.

As we explore the integration of AI into our daily lives, we can observe its application in email services, where predictive text prompts streamline sentence completion through a simple swipe.

But what is so revolutionary about AI, and how has it progressed over the last century up to present time?

A Brief History of Artificial Intelligence


Artificial Intelligence in AEC: Advancements in the Industry

The AEC sector, an embodiment of the teamwork of architects, engineers and construction professionals, plays a crucial role in both local and global economies, contributing an average of 6% to the global GDP.

In a 2018 McKinsey & Company assessment of AI adoption across 13 industries, construction ranked among the bottom three for current AI use. Despite this, a study published in May 2023 highlights a rapid evolution in the AEC tech industry, characterized by heightened investments and a surge in startup numbers. Around $50 billion flowed into AEC tech from 2020 to 2022, marking an 85 percent increase compared to the previous three years.

A brief history on Technology in the AEC industry


Undoubtedly, AI is significantly transforming various industries. Experiencing a notable rise in technology adoption over the past three years, the AEC industry has seen the influence of this trend spreading across various areas:



Finally, buildings are evolving beyond mere walls, roofs, and masonry. With the integration of Internet of Things (IoT) devices, sensors, and automation systems, such as Para’s digital twin and building analytics platform, buildings are becoming intelligent structures. However, “intelligence” stems from data analytics, as data without analytics is comparable to a puzzle without assembling—a collection of pieces waiting to reveal the complete picture.

Here are some of the transformative ways AI has transformed the AEC industry.

Unleashing Reinforcement Learning for Optimal Energy Solutions

The prevailing energy consumption optimization approach is primarily reactive, responding to unfolding events without proactive planning. Using tools like Reinforcement Learning (RL) becomes crucial to transition to a more strategic stance.

RL is a branch of AI concerned with the decision-making process. Likened to playing chess against a computer, the agent interacts with the environment (chessboard and PC), with each action resulting in different states and earning rewards. RL aims to maximize rewards by guiding the agent to learn optimal steps for efficient decision-making.

In the context of the AEC industry, specifically for energy consumption optimization, RL is being explored to enhance processes such as scheduling HVAC systems.

Taking the example of HVAC optimization, the digital facility manager, acting as the agent, controls the Building Management System (BMS) impacting the HVAC zone—the designated environment. Unlike electrical systems, HVAC systems require advance planning, and hasty reactions may have detrimental consequences due to the dynamic nature of the environment. To mitigate risks, a digital twin of the environment is created, allowing the RL agent to learn and optimize without immediate real-world consequences.

The effectiveness of RL depends on the user. While one may possess extensive knowledge in this field, it is essential to apply it judiciously, avoiding trivial scenarios like the chess example. Recognizing the strategic use of these skills is crucial, reflecting the forward-thinking approach embraced by Para.

Fault Detection & Diagnosis: Mitigated Downtimes & Increased Savings

Mitigating unscheduled downtime is crucial to reducing operational expenditure, extending machinery lifespan, and minimizing costs and risks. However, pre-defined engineering rule-based solutions have limitations in detecting faults not previously accounted for by these rules.

To address these limitations, machine learning models can be used to learn the complex relationship between the inputs by using the machinery’s historical data. Two paradigms of machine learning – Supervised and Unsupervised learning – can be used to refine the model by minimizing deviations between actual and predicted values, or by learning the underlying relationships between inputs. Any deviation in the machinery’s behavior can be detected, and these deviations can be investigated by subject matter experts to prevent major issues.

During the investigation phase, AI, particularly clustering, proves highly beneficial in categorizing faults with similar patterns. Ultimately, by linking the outputs of multiple AI systems together, an advanced classification system that predicts the fault label associated with the root cause of the issue can be achieved.

Incorporating AI into the fault detection system ensures proactive maintenance, minimizing downtime, and reducing operational costs.

The importance of AI applications in bringing about positive changes.

The wealth of available data surpasses the realms of IoT and building systems, and more extensive source lies elsewhere. From blueprints to material specs, there’s a digital record for almost everything. All types of data sources are serving as the driving force for numerous emerging AI applications:

AI-powered tools can analyze vast amounts of data and generate intelligent design suggestions, thus optimizing the design and planning processes, ultimately resulting in more efficient and sustainable structures.

LLM-powered tools are on the rise, and engineering tools infused with LLM capabilities are beginning to emerge, empowering users to access information, query databases and offering numerous functionalities beyond our current imagination such as converting text prompts to concept designs.

AI and sensor-enabled wearables hold the promise of revolutionizing the construction industry, offering workers instantaneous information about their environment and their well-being (i.e. fatigue and stress monitoring), providing them enhanced safety and productivity.

AI-powered drones and robots are set to revolutionize the AEC industry, bringing enhanced efficiency to site surveying, inspections and construction operations. The synergy of drones, robots, and AI, provides construction automation thus enhancing productivity, ensuring safe operation in high-risk areas, and meeting deadlines.

The future of AI in AEC

With the expanding use of AI-powered tools, the AEC sector’s productivity surge, estimated by McKinsey, could translate into a market value of approximately $1.6 trillion. The rapid adoption of AI brings challenges, such as addressing data privacy, maintaining ethical practices, and overcoming expertise shortages. Prioritizing workforce upskilling and reskilling, along with developing applications that balance customer experience and workflow execution, are essential.

Ultimately, in the face of challenges such as climate change and urbanization, AI emerges as a transformative force, fostering efficiency, sustainability, innovation, and safety.