A pilot study of a 3D game environment for construction safety education

Abstract

Education is crucial to promoting a safe and healthful working environment in construction and teaching students to identify potential job hazards is its very first step. The authors proposed a 3D video game, Safety Inspector, to provide a comprehensive safety training environment in which students assume the roles of safety inspectors and walk the game site to identify potential hazards. Violations of different modelling difficulties and various levels of safety knowledge for identifying these violations were listed to guide the development. The game was also designed with an agenda to address features such as realism, self-learning, non-linearity, interactivity, etc. Torque 3D game engine was leveraged for implementing the game system and Autodesk 3ds Max as well as MilkShape 3D were used to create the needed but unavailable 3D objects. An important development strategy was approximation - this applied to both the definition of object collision boundaries and the texture mappings. A small group of students from the Department of Construction Management at the University of Washington helped test the preliminary game system. The testing results indicated that with the game students increased their learning interests, enjoyed the learning process, and were motivated to refresh their safety knowledge. In addition, students also showed optimistic attitudes towards using the game scoring as a way to reflect their safety knowledge. In overall, the evaluation results suggested a positive outlook of the game and encouraged the continuous development of Safety Inspector. However, the prototype system did not incorporate all the desired violations or features and should be further enriched in its next version. Pedagogical issues newly discovered during the game evaluation process are to be addressed as a part of the future work.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 118 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.