How do design decisions affect the climate, costs and construction time of bridges?
“We can simulate thousands of bridge alternatives, and get answers in seconds,” Alexander says. The goal is to help teams see how choices made up front affect climate impact, cost and build time – and to do it early enough to guide decisions when it matters most.
Alexander is a bridge engineer at Skanska Teknik in Gothenburg, Sweden. Since early 2024, he has also been an industrial PhD student focused on how machine learning can help optimize bridge structures. His aim is not to replace engineering expertise, but to give colleagues better support when making decisions in complex projects. “I want to provide models that show the consequences of our choices early on – whether that’s climate impact, cost or buildability – so we can decide based on evidence, not just experience,” he says.
In partnership between Skanska and Chalmers University of Technology, he has defined a research area closely aligned with both the company’s needs and his own interests. He has long been curious about research, felt the timing was right, and took part in writing the application, shaping the topic together with Skanska – a process he found both fun and rewarding.
At the core of Alexander’s work is synthetic data, generated from parameters and variations across bridge types. By training AI models on this data, the team can answer questions in seconds that previously required extensive manual calculation. They can simulate thousands of solutions and see how a particular choice affects carbon emissions, construction time, material costs, transport volumes or even the size of the work area.
Over time, this research can help reduce waste and strengthen competitiveness, especially in the tendering phase. A broader aim is to free engineers from repetitive tasks so they can focus on creative problem‑solving, delivering better solutions for both the climate and customers, often taxpayers.
Creating change
Alexander has been doing his research for 18 months, and the next major milestone is his licentiate thesis (a Swedish exam similar to half-way review), including two conference articles and one in a peer-reviewed journal. A sneak peek at work so far was presented at a conference in Japan in May 2025.
Looking a few years ahead, he will summarize the results in a doctoral dissertation. Above all, he hopes the work will drive change at Skanska and across the industry, resulting in tools that can be used widely and make a real difference.
He describes the project as highly rewarding, with strong support from both Chalmers and Skanska, and finds the combination of research and practical work incredibly instructive.
Alexander Kjellgren: “We can simulate thousands of bridge alternatives, and get answers in a few seconds.”
Related articles
