English Articles

Construction sites could be fully automated with AI

Construction sites could be fully automated with AI
© Sven Mieke | Unsplash

[KIT] – How can artificial intelligence (AI) support persons working in the construction sector? Which tasks can be learned by machines? These questions were addressed by the KIT-coordinated project “SDaC – Smart Design and Construction.”

“We have developed a platform, on which project documents, such as delivery notes in PDF format, can be digitized and structured,” says the Scientific Head of SDaC, Professor Shervin Haghsheno from KIT’s Institute for Technology and Management in Construction (TMB). The SDaC project partners developed nine AI demonstrators to support organizations of the construction sector in the planning and implementation of construction projects. “We attached high importance to transparency and explainability,” Haghsheno adds. The demonstrators are now available on the platform and can be tested to experience the added value of AI in the construction sector. Although the project has been completed, TMB will continue to operate the platform and the demonstrators in order to support the exchange of knowledge and experience relating to AI in the construction sector.

“The findings obtained in the course of the project are very important to our research on digitization the construction sector and human-machine collaboration,” Shervin Haghsheno says. According to Project Head Svenja Lauble, TMB, the construction sector is a strongly fragmented ecosystem with a large number of interfaces. “That is why loss-free transmission of information among organizations and software systems is very difficult. With SDaC, we have demonstrated the added value of AI, in particular for small and medium-sized companies. They often have to handle numerous data formats that are neither digital nor structured.”

Platform Enables Digitization of Documents
The SDaC project partners also compiled a list of 230 software companies offering AI for the construction sector. On a website, AI experts can offer their services to construction and construction software companies. In addition, training courses and networking events are organized. Based on the demonstrator “Lieferscheine digitalisieren” (digitizing delivery notes), researchers have developed an app that will be available in the app stores for download in fall 2023. Every construction site receives delivery notes in paper format which have to be processed manually for auditing and documentation purposes. The app can be used to digitize them.

Follow-up Project on Smart Remediation Measures
Some of the demonstrators developed for SDaC will be optimized under the follow-up project “NaiS – Nachhaltige intelligente Sanierungsmaßnahmen” (sustainable smart remediation measures) that is also coordinated by KIT. NaiS is funded by the Federal Ministry for Economic Affairs and Climate Action (BMWK) and will link data from different sources on a digital platform with the help of AI technologies. The goal is to objectively assess and optimize remediation measures.

The startup Valoon in Dortmund also is a result of the SDaC project. Valoon focuses on the smart processing and structuring of information from existing communication channels, such as WhatsApp.

More than 40 partners from science and industry were involved in SDaC. The project funded by BMWK with around EUR 9 million for a duration of 3.5 years won a Germany-wide AI innovation competition.

More Information on the Project SDaC – Smart Design and Construction