Digital twins are opening new doors for the architecture, engineering, and construction (AEC) industries.

Previously, bridging the gap between imagination and actual building was a laborious and time-consuming process. However, with the advent of digital twins, collaboration among stakeholders and across geographical locations has become more streamlined.

Nevertheless, builders have encountered certain obstacles that could impede the widespread adoption of digital twins, especially for those who have yet to try it.

To hammer out some solutions, experts gathered for a forum titled “Building New Worlds in Real Time,” organised by Jicara Media, and hosted by Dell Technologies and PTC System.

Towering issues

With the advent of digital twins, organisations are now able to collect data at an unprecedented scale.

While this development is crucial to unlocking actionable intelligence for the AEC industries, many businesses are currently struggling with the transition from data collection to meaningful insight, noted Derek Herbert, Chief Information Officer of DP Architects.

“The issue lies in taking the time to analyse the data, or to reuse the data to make it more effective. It is only when we dedicate sufficient resources and time to harvest the data, and then make sense of the data that we can truly produce better outcomes,” he said.

Herbert used ChatGPT as an example, which can be helpful in various ways, but still has a long way to go in providing sufficient real information.

“At the end of the day, you still have to train the bots and clean the data before you can actually get the right answer,” he continued.

Meanwhile, prior to building a digital twin, a collaborative effort among project stakeholders and a significant amount of groundwork need to be done to ensure the model’s accuracy, remarked Daniels Chandra, Director, Digital Technology, ONG&ONG Group.

“As consultants, we must consider the wealth of knowledge and preparations needed to achieve the desired end goal, such as determining appropriate design outcomes based on the client requirements and compliance with building codes. It entails organising numerous data containers and making critical decisions about which data to include,” Chandra said.

Then again, a shift in mindset is necessary to ensure that all teams working on a project remain on the same page.

“Having been trained as an architect, I went through the process of creating the design manually using traditional pen-and-paper methods. Then, CAD (computer-aided design) came along to digitise the pen-and-paper process, followed by BIM (building information modelling) software that focuses on the ‘I’ (information, data-driven, and object-oriented processes). The transition from pen and paper to BIM on a computer represents a significant shift in mindset,” Chandra mused.

He explained that in the days when the industry relied on pen and paper or 2D CAD, it was sometimes difficult to identify mistakes or gaps. “However, with the introduction of BIM, we can now design, simulate multiple design scenarios, and verify them within the BIM platform before the actual construction on site,” Chandra added.

Finding the right fit

DP Architects has been using an HPC platform for quite some time now. However, Derek Herbert, the firm’s CIO, has observed a noticeable disparity among different user groups when it comes to adopting the technology.

“It is crucial to involve the right stakeholders in a new platform like Omniverse. This is where maturity plays a role. While some users demonstrate proficiency, we have witnessed certain clients encountering challenges. If the appropriate stakeholders are not engaged, attempting to collaborate using Omniverse becomes limited in its effectiveness,” he said.

Since DP Architects runs its own high-performance computing (HPC) cluster, Herbert has experienced instances when daily operations slow down considerably due to unforeseen issues.

“With the constant adjustments in mass study or form study, and then running through the models again, things can slow down. In the meantime, clients are eager to see what the building is going to look like. This compels us to rethink the computational demands placed on our HPC system. The resource allocation depends on the scale of the project we’re handling,” the CIO shared.

Storage requirements

Another consideration when embarking on virtual twin projects is data storage.

In the case of ONG&ONG Group, all workloads and applications are hosted 100% on the cloud, which aligns with the company’s data strategy and its exploration of AI use cases.

“We have partnered with a cloud service provider that is compatible with our design and collaboration tools. This allows us to author, edit, coordinate, share, and store everything in the cloud, offering numerous benefits. However, there are pros and cons to consider. For instance, if there is no internet connection, immediate file access becomes limited,” ONG&ONG Group’s Daniels Chandra said.

Despite this limitation, adopting cloud storage has enabled the firm to share the workload across all its offices, facilitating seamless collaboration with project stakeholders and supporting remote work. This has proven especially integral for business continuity during the pandemic.

In contrast, DP Architects’ Derek Herbert clarified that HPC is not as affordable as commonly perceived.

“While efforts are being made to make the technology more affordable, I still believe it’s a significant investment. Implementing HPC should be approached as a substantial commitment for those choosing this path. The typical lifespan of an HPC cluster is around three to four years due to its constant 24/7 operation. We run a hybrid environment, with equipment on premises and in the cloud. However, certain large data sets may never be migrated to the cloud due to their size and the associated costs, whether it’s data transfer or storage. Therefore, we need to be mindful of costs,” Herbert explained.

A dent on creativity?

As AI continues to make significant advancements in the AEC industries, particularly in image generation, concerns arise about the potential stagnation of creativity in an era dominated by templates.

Daniels Chandra from ONG&ONG Group recognised the role of AI in the ideation process and highlighted its time-saving benefits, similar to HPC.

“AI can help create and enhance some photorealistic design intent, mood boards, and supporting design elements based on the original architect’s hand sketches or simple BIM model. It offers architects and engineers a multitude of options and possibilities, but it does not necessarily replace their roles,” he said.

On the other hand, Derek Herbert from DP Architects foresees AI playing a larger role in the production phase.

“We haven’t reached the point of incorporating AI into the ideation or concept stages yet. It’s pretty challenging to replace the human mind, as far as originality is concerned,” he stressed.

Moving forward, both experts shared a set of recommended steps to maximise the benefits of virtual twin technology:

• Adopt a data-driven and information exchange mindset while addressing interoperability issues across multiple tools used in a project.
  
• Establish a common data environment for the project stakeholders to embrace a single-source-of-truth approach and enhance data security.
  
• Introduce collaborative practices towards the common goal within the AEC industry.

The adoption of digital twins in the AEC industries has the potential to change the approach to design, construction, and project collaboration. To fully harness the technology’s potential, it is crucial to address challenges related to data security, streamline workflows, and foster innovation.

By prioritising these aspects and harnessing human creativity, digital twins can drive innovation in designing and constructing physical spaces. The possibilities and applications of digital twins are vast, holding immense potential to transform the industry.