Nuclear Radiation Safety - Protecting staff in radioactive environments

The brief:

This project was driven by Innovate UK and the Nuclear Decommissioning Authority with the aim of designing and developing a virtual reality simulation (VRS) training tool, which can be used by nuclear industry operatives to train in procedures and processes relating to the use of personal protective equipment (PPE). This followed feedback from Sellafield Ltd that safety concerns exist in this area.

“The purpose of this is to ensure that these procedures are effective and safe before performing them in ‘the real-world'”

The VRS training tool needed to be an immersive, high fidelity, virtual environment that would enable trainees to practise and rehearse procedures relating to PPE, in particular the use of air-fed suits. The purpose of this is to ensure that these procedures are effective and safe before performing them in ‘the real-world’, thereby minimising the risk to operators. Further to this the VRS training tool was to coach appropriate hazard perception behaviour.

This proof-of-concept project was led by a team of academics and industry partners with extensive experience of virtual reality simulation, the psychology of learning, and a detailed understanding of the technical processes associated with Nuclear Decommissioning.

What we did:

In this first proof-of-concept phase of the project, we built and tested a working prototype of the VRS training tool using real-time gaming design software.

The starting point for VRS development was Operational Experience Feedback (OEF) gained from research inside the  Sellafield and Magnox COOL database. The bulk of safety incidents were related to failures to adhere to the correct dressing procedure, failures to identify suit damage, and failures to identify hazards in the work area.

“We developed a virtual dressing room environment and a virtual air fed suit, which enables trainees to practise and rehearse procedures relating to PPE”

Following advice from safety experts and following industry procedure, we developed a virtual dressing room environment and a virtual air fed suit, which enables trainees to practise and rehearse procedures relating to PPE.

Operators are tasked with following instructions and correct procedure to learn how to appropriately inspect the suit. A suit wearing simulation and a hazard perception task were developed; the simulation emulates the wearing of a suit with visual and auditory effects. Work place hazards include hose obstructions, sharp edges, poor visibility and increased noise.

The research:

We arranged a testing session at the Training Centre of Magnox Ltd based at Oldbury Power Station. As well as experienced operators, the VRS was tested on members of the Health Physics Team. The simulation included a task in the final hazard perception phase that required users to search the area for markers.  The purpose of this task was to direct the user into all parts of the work area where they would find a range of hazards and experience variable environmental factors. The task was  designed to emulate an area survey for radiation classification purposes. The VRS is instrumented with live data gathering tools that gives the trainer a timestamp and event log information output, in order to evaluate user performance.

Using these data, we adopted a validation approach we have used for other VR training projects.

The results:

Results provide initial proof-of-concept that the VRS is considered an effective training aid. Feedback suggested that ‘the VRS would be better than current classroom training’; that ‘the realism was surprising and quite convincing’; and that users’ thought ‘it would be really useful, as it allows us to plan out work and think through where the hazards might be’.

“More experienced workers were able to complete the task in less time”

Objective data captured during in-task performance (analytics from the VRSv1) were subjected to statistical analysis to explore the relationship between role experience, and VRS task performance (‘construct validity’). While the correlation equation did not reach significance, the line of best fit indicated a trend that more experienced workers were able to complete the task in less time. This will be explored further in Phase 2 with a larger sample size, and improved measurement. The success of Phase 1 highlights the technical feasibility of VRSv2.

The VRS training tool provides, (1) improved training outcomes for learners, ultimately leading to improved safety outcomes; (2) more efficient training delivery, placing a lower time burden on personnel and resources; and (3) a more cost effective and sustainable form of training.

The project has since moved into a second phase, where the training tool will move from prototype phase into creating a working, fully functioning simulation tool. This simulation will feature more detail in the dressing procedures, enhanced analytics and improved hazard identification stages.