Strategies for construction hazard recognition
Phase 1: Development of SAVES
Download 2.75 Mb. View original pdf
|
- Navigate this page:
- Development process
| Phase 1: Development of SAVES The objective of the first phase was to develop a highly reliable tool to train construction workers to more effectively detect hazard stimuli in dynamic construction environments. In order to achieve this objective we developed a high fidelity System for Augmented Virtuality Environment Safety (SAVES) platform that integrated effective principles from modern theories of learning. SAVES was designed to immerse trainees in a hyper-realistic, D environment to provide a basis for contextual and self-directed learning (Dolmans et al. 2005). In this environment, workers are exposed to diverse but representative work scenarios that have been recently encountered on projects in the US. The tool also includes functionality to assess worker hazard recognition 47 performance, provide feedback based, and equip workers with a holistic approach for hazard stimuli detection using retrieval mnemonics. From a training design perspective, this was based on our previous work with Generation Y trainees from a large logistics company. Design criteria included immediate feedback, fun, face valid, crew level interaction. Development process SAVES was developed through the integration of an actual Building Information Model (BIM) of a petrochemical facility with a large database of construction scenarios images using the Unreal Development Kit (UDK) gaming engine (See Figure 1). The BIM model was from a recently completed project provided by a research team member. The construction images – about 700 images – were also gathered from members of the research team. A subset of images with different hazard types (e.g. falls, struck by, exposure to harmful substances, etc) and varying severity levels were selected by the expert panel for integration into SAVES. Hazards associated with this subset of photographs were identified by the expert panel through brainstorming sessions, which was essential to provide active feedback to the workers during the virtual learning experience. The BIM model and the construction scenario images were imported into the UDK environment, and a set of training scenarios were designed using ActionScript and UnrealScript. As part of SAVES, a reliable and interactive interface was developed. The interface has two key components (1) The user-controlled Hint system which provides Hint in D environment to minimize the search time for locating hazards for the users – programmed in C+ and UnrealScript, and (2) The Scoring system which provides real-time feedback to users as they conduct hazard recognition programmed in C+ and UnrealScript as well. |