If you attended Digital Construction Week 2024, you'll know that Artificial Intelligence (AI) was a major theme of the two-day event. Karoliina Torttila, Director of AI at Trimble, led a conference on the role of AI in the renovation industry. Missed it? Find out in this Q&A session some of her thoughts on AI and the value it brings to the built environment...
Q) What is AI and how can we use it in the built environment?
I think the implications of AI for the built environment are so varied that it's impossible to give a single answer to this question. That said, there are two benefits that are perhaps the most exciting for me. The first is the role of AI in creation, which allows more freedom in our design processes. This could be the ability to design from different data modalities, from your floor plans and scans of the physical world to simple natural language.
The second is unrivalled efficiency in the creation of deliverables - whether it's taking your project specifications, creating your submission logs, then matching the two; automating costing; or processing your reality capture data.
Q) How are we already using AI?
Today, we're able to efficiently capture a lot of data about the physical world. However, doing so creates a volume of data that we simply can't process manually. AI can help by taking this data and processing it in real time with a high level of accuracy, to turn this data into easily actionable information - in essence, helping to connect the physical and the digital.
Q) What are the advantages of AI for designers?
Firstly, there are the rapid iterations that AI enables in the design phase. Then there are the efficiency gains in the design of details and complex structures such as reinforcement. So, even in the design phase, AI plays a valuable role, both in exploring alternative designs and in creating highly detailed ones.
Q) How can we use AI to transfer data from the design office to the site?
Thanks to AI and other technologies, we have already succeeded in transferring design data from the office to the construction site. Now, we can look forward to bringing that data back from the construction site to the design. Whether it's a worker filling in inspection reports, or tracking site progress to automatically trigger payments, a two-way flow of data needs to take place. AI is an important enabler for achieving this on a large scale.
Q) Where do you see AI solving some of the industry's current problems?
In recent months and years, AI assistance has been the subject of various discussions. For me, AI assistance can be thought of as the smart friend in the room, who can hold your hand while you hone your skills. As we see more and more of our industry's "veterans" retire and we start to lack mentorship, I think AI has a role to play in helping to mentor young professionals and accelerate their learning curve.
Q) How can we implement an AI strategy, and what does this mean for our data?
Simply put, it's impossible to implement and deploy AI without first having a solid data strategy and bringing that data together in a common data environment. These fundamental elements need to be in place before AI can really be leveraged on a large scale.
Q) What are the challenges and impacts of AI in industry?
While AI offers us unparalleled automation, it also presents the risk of moving too fast and failing to implement the necessary checkpoints. I think it's the responsibility of all of us to understand how we can take advantage of this automation and implement the necessary control and verification processes, so that we retain confidence in the results and deliverables.
Q) Are you an AI optimist or pessimist?
I'm definitely more on the optimistic side. Personally, I'm a big believer in open-source data and these transparent discussions about the evolution of this technology. No one - not a technology provider, not a contractor, not a design consultancy - acts alone. It's through conversations as an industry that we can ensure we move forward collectively, logically and with confidence.
Find out more about AI at Trimble: https: //www.trimble.com/en/ai-at-trimble
