Business Finland has awarded €20 million in co-innovation funding to a project focused on productivity challenges in architectural services engineering. University of Tampere Coordinating a consortium of companies working on the research component and practical deployment.
The initiative, called AI Champion, will bring together researchers and companies to investigate why architectural services engineering projects run into delays and cost overruns, and whether “human-like” AI agents can reduce the friction of data handoffs throughout the supply chain.
Business Finland is Finland’s public innovation funding organization. Tampere University is a Finnish research university and will coordinate the research activities of the project with a research budget of approximately 5 million euros.
Business Finland funding targets supply chain data gaps slowing down projects
This project focuses on recurring operational issues in building services engineering, such as fragmented data, inconsistent formats, and handoffs that don’t transfer cleanly between parties.
Pia Solmunen, associate professor and project coordinator at Tampere University, said the project had been in preparation for 20 months and was linked to Finland’s broader data economy challenges.
“The application process and preparation for this unique AI project, which will revolutionize the fields of architectural services engineering and civil engineering, took 20 months,” said Solmunen. “We are very happy that Tampere University has secured funding that will enable us to conduct interdisciplinary research and build an AI hub in Finland. This project is one of the most important data economy pilot projects of the Finnish Ministry of Economy and Employment.”
Solmunen also points out that basic formatting and workflow issues cause time loss and rework, saying, “Design documents are typically always in PDF format, which the contractor prints at the construction site. The big question here is why the data is not transferred electronically from one party to the next, usually creating an information gap between the parties at some point in the process.”
Tampere University plans “GPT Lab” workflow and 100 AI agents
AI Champion will develop 100 AI agents aimed at automating and improving information flow in building service engineering supply chains. The project description frames these as “human-like” agents designed to support processes and operational models when information bottlenecks are identified.
In the model we have outlined, when a disruption in the information flow is detected, the information is routed to the GPT Lab, called the Virtual AI Lab at the University of Tampere. Researchers then combine relevant information to build agents that can support specific process steps.
Solmnen describes the expected outcome as a change in the way work is organized, with some tasks removed and new ones introduced as workflows become more data-driven.
“Certain work steps will disappear in the process, but at the same time new work steps will be created. More data-driven and AI-controlled processes and services will be developed during the project, increasing the competitiveness of the consortium companies nationally and internationally. Moreover, Finnish companies focusing on artificial intelligence will be able to find new business opportunities in the civil engineering industry,” said Solmunen.
A consortium of construction and engineering companies
The project consortium focuses on testing new agents on real corporate projects, covering multiple stages of the architectural services engineering lifecycle from design to maintenance, rather than maintaining development in a lab environment.
Tampere University will contribute to an interdisciplinary team spanning the built environment, management and business, information technology and communication sciences, with a focus on construction workflows, information management and AI agent development in the context of building services engineering.
The University of Oulu offers expertise in product management and business digitalization under Professor Harri Haapasaro. Professor Harry Harpasaro defines “data flow” as something that is continuous, automatic, and available in real time without manual workarounds.
Harpasaro said: “This project is very important and includes many interesting stages. Our main expectation is that we will be able to achieve a genuine data flow. Data flow is not just the transfer of data from one system to another. Our expectation is that data will flow continuously and automatically through various processes and systems in the appropriate format and be available in real time without manual intermediate processing. Once we can link to that data, we can start talking about rich product data and big opportunities for the future.”
The project is scheduled to start at the end of this year and run until June 2028.
