AIAKI: Artificial intelligence to manage AIA in BIM
Subproject:
Development of an ontology-based AI system for AIA configuration.
Project Duration: 01.09.2022 – 31.08.2024
Budget: ca. 700 k Euro (total); ca. 229 k Euro (TUD)
Objectives
The AIA-KI project is divided into 6 work packages (AP). The work packages are processed in three project phases. In the first six months, the development of strategies and design documents takes place. This conceptualization and planning phase is concluded by milestone A.
The development phase comprises a total of three steps, of which the first two steps serve to develop the components of the software platform and the hybrid AI system (AP B and C). Milestone B is reached when the compilation and formalization of the knowledge structure and configuration rules as well as the hybrid AI system (AP D) are completed.
In the final project phase (AP E), the developed components are integrated, and the overall system is tested. Thus, Milestone C represents the end of the project and the completion of a functional AI-based AIA configuration platform. The activities of AP F are for project and risk management as well as public relations and the preparation of the exploitation plan. These tasks will be worked on continuously, accompanying the project.
The TU Dresden participates in the joint project AIA-KI under the leadership of AEC3 GmbH together with GI-Consult GmbH. The contribution of the TUD focuses on the development of a hybrid AI system for the semi-automatic creation of AIA documents (AP C, Phase 2). Before that, TUD participates in the requirements analysis and concept development, especially in the object design (AP A - Phase 1). Furthermore, in Phase 2, TUD participates in the integration of the hybrid AI system into the AIA platform (AP D). In project phase 3, TUD participates in the validation of the platform and optimizes the interfaces between the AI system and the AIA platform. TUD is continuously involved in project and risk management tasks.
AIAKI website | AIAKI Flyer | internal |