Together with other 11 project partners, PMG has been working on the ESKIMO research project since April 2020. The aim of this project is to apply artificial intelligence (AI) methods for the realization of a far more efficient construction management. AI is to support the project participants of a construction project in a technical as well as commercial quality assurance and to create a basis for intelligent construction logistics.
Based on the objectives, three pilot modules will be developed in the course of the project, which in the future should help to keep a close eye on the construction site with a significantly smaller investment of resources and time.
The commercial quality assurance is about the comparison between the actual state on the construction site and the target state of the BIM model. This reconciliation must primarily create a partial automation of the release processes for invoice verification runs and thus an increased efficiency in the project process.
By using AI methods, the degree of automation of technical and commercial construction supervision is increased significantly.
PMG is developing the pilot module “Commercial Quality Assurance” based on three scenarios:
1) permanently up-to-date room book with all available data sheets;
2) partial automation of invoice verification;
3) automatic building status determination.
Up-to-date room book with all available data sheets
The goal of scenario 1 is to have a room book that is always up-to-date with all available data sheets in the project, which fills itself from various sources and updates itself automatically (incl. assignment of data sheets to rooms/areas with additional information regarding trade and subcontractor).
Partial automation of invoice verification
The goal of scenario 2 is to partially automate invoice verification so that all data required to verify an invoice is available online to the releaser directly during invoice verification/release.
Automated building status determination
Automated building status determination is based on the BIM model and reality. Using regular as-is images of the construction site, the time of installation of the individual construction elements can be determined and this can be compared with the planning based on the model. In this way, the performance status can be continuously monitored and meaningful economic key figures derived.