Is AI still used in civil engineering?

More security thanks to artificial intelligence

The research project “Smart Design and Construction” (SDaC), in which 40 partners from the construction industry are participating, aims to optimize safety on construction sites with the help of artificial intelligence. In various applications, inter alia. go through the partially automated creation of risk assessments as well as the partially automated recognition of hazards when placing an order.


SDaC has set itself the interaction between humans and technology as a central task. Eight use cases will be designed, tested and implemented during the three-year research project. At the core is always the user of the application - thanks to the "Design Thinking" method, the users are involved in the development of the applications from the start. In this way, the research project ensures that the applications are developed in accordance with the requirements of the construction projects.


Use case risk assessment

Together with BG BAU, SDaC deals with risk assessments for construction sites in one application. The focus is on two main aspects: on the one hand, the creation of risk assessments before the start of construction and, on the other hand, the data-based detection of hazards during the construction phase.

Basically, a solution is to be created that relieves employees of the tiresome work involved in creating and tracking risk assessments and enables more expert knowledge and individual consideration of hazards on construction sites.


Semi-automated creation of risk assessments

In the first step, the building contractor makes the existing data available on the SDaC platform. In the construction industry, these data are typically available as GAEB, Excel, PDF and CAD files. The AI ​​algorithms provided by the SDaC platform analyze the data supplied and make a proposal for the project-specific risk assessment to the entrepreneur, site manager or foreman. In practice, the process looks like this, for example:

  1. GAEB file is read.
  2. The algorithm examines the service texts.
  3. The algorithm makes suggestions for the risk assessment based on the service description. Example: A sewer construction to a depth of 5 m is advertised, the algorithm recognizes from the depth that a sheeting is necessary. The wizard lists the relevant hazards that the user still needs to confirm or change in details.
  4. After adjustment by the user, the user can choose the format in which the construction site information is to be made available. It can, for. For example, the PDF format can also be selected as the output.

This process shows schematically the goals of SDaC. The usual data sets available in construction companies are to be used, after which added value is to be created with the help of AI applications and the information generated is again provided in a format that is usual in construction.