Mon, 16 December, 2024
The SafeXtend project is addressing ongoing challenges in relation to skills shortages and the rising cost of training, as well as the safety of site personnel in the construction industry, namely the multitude of professional and personal skills required: from health and safety management, and operation of equipment, through to problem-solving, team working and effective communication.
Current industrial training methods include traditional, classroom-based courses and off-the-shelf training packages, both of which do not realistically simulate the hazards that can potentially occur on building sites, and limit their usefulness to construction contractors and operatives alike. For example, in-person classes can be unfeasible due to the need to schedule training around fluctuating project timelines, while lack of reliable internet access on job sites mitigates against online course participation. Research has also shown that course content does not always adequately address contractors’ needs in relation to specific skills, processes and equipment required on job sites, and therefore lacks relevance. As a result, attendance on, and engagement with, construction industry training courses is generally low.
By developing an immersive, virtual reality (VR) training platform powered by artificial intelligence (AI) that provides realistic 3D modelling of building sites, SafeXtend offers a solution to these needs. It incorporates specialised scenarios, for example health and safety, and equipment operation, tailored to different construction sectors such as infrastructure and heavy civil engineering. These enable construction personnel to practice communication, problem-solving and team working through multi-user VR functionality, and AI techniques including computer vision and natural language processing provide personalised feedback to individuals on their performance.
The platform's AI capabilities will automatically convert 3D site scans into interactive digital shadows and generate customised training programmes for each scan/model. This will immerse trainees in realistic jobsite simulations via scenario-based VR environments that have been targeted to their particular needs through embedded skills assessments. Use of conversational AI, avatars and different scenarios will help improve trainees’ engagement, for example using visualised hazards rather than static checklists. Captured proficiency data will provide detailed reporting that can be used to gain insights, such as benchmarking training programmes against return on investment over time.
The immersive nature of the platform’s environment will help trainees to remember what they have learnt, leading to improved productivity, and the potential for injury while learning is lessened because programme participants are trained in a way that reduces risk. Additionally, the overall cost of delivering training in this way will be much lower than using traditional methods due to the platform’s scalability.