The growing use of AI in construction has created a need for datasets that can support reliable model development, evaluation, and reuse. However, construction datasets vary widely in quality and characteristics, making it difficult to know when they are suitable for AI tasks.
To address this, the paper proposes a multidimensional operational framework for assessing the AI readiness of construction datasets. The framework is designed to help researchers and practitioners systematically evaluate datasets before using them in AI workflows.
The emphasis on evaluation and reuse suggests the framework could support more robust benchmarking and sharing of datasets across projects. The authors position the work as a response to the rapid expansion of AI applications in the sector.