Construction Company and Province of Overijssel Test ChatGPT

A construction company in Overijssel recently conducted a trial to investigate how artificial intelligence (AI) can accelerate work on the construction site. The focus of this practical trial was on tasks such as planning, quality control, and reporting. The aim is to reduce errors, lower costs, and create a safer working environment. The trial also takes GDPR and the European AI regulation into account, especially given the public nature of many contracts.

The initial results of the pilots indicate that algorithms are particularly useful for repetitive tasks. Examples include organizing photos and notes or automatically summarizing long emails, which helps to reduce search time and manual work. Although generative AI can assist in converting construction reports into concise action points, human oversight remains essential for final decisions. Image recognition can also provide support in quality control by identifying defects, although rechecking by a foreman or quality inspector is always necessary.

The greatest benefit of AI emerges when it integrates with existing software like ERP and BIM. This prevents duplicate entries and ensures streamlined processes. It is crucial that data is of high quality, as incomplete information affects the results. The principle of "garbage in, garbage out" applies; therefore, it is essential to standardize and clean data. Moreover, privacy regulations such as the GDPR and the new European AI regulation play an important role, especially regarding the use of image recognition on the construction site.

Implementing AI requires change management and a careful approach. Training teams, clarifying responsibilities, and appointing a product owner are necessary for a successful transition from the trial to practice. When procuring AI solutions, it is important to pay attention to data location within the EU, security, and exit options. Collaboration throughout the entire chain, from subcontractors to clients, is crucial. By contractually formalizing necessity and usage rights, scaling AI applications can be made easier and more transparent.

Read the full article from AI Insider.