Speaker
Description
Abstract:
Reading English construction drawings asks students to work with words, symbols, dimensions and spatial layouts all at once, and many Vietnamese civil engineering students struggle at this point. Generative AI tools now give them a new kind of support, but we know little about how students use these tools alongside the scaffolds their teachers design. This qualitative study explores how 22 second-year civil engineering students at a Vietnamese university use AI tools when reading English construction drawings, and how they see AI in relation to teacher-designed multimodal translanguaging scaffolds. Data were collected through semi-structured interviews and reflective journals and analysed thematically. Students described AI mainly as a fast translator and term explainer, and they moved between English and Vietnamese to check meanings. They also noted that AI often missed the visual and spatial logic of drawings, such as symbols, section views and dimension chains. For these, they still turned to teacher scaffolds. The findings suggest that AI works best as a complement to well-designed multimodal scaffolds rather than a replacement for them. They also point to a need for guided AI use in engineering ESP classrooms.