(424b) Utilizing Augmented Reality (AR) and Standard Alarm Management Practices to Provide Real-Time Information to Refinery Workers in High-Stress Environments: Proof of Concept

In the event of an incident at a refinery (accident, severe weather, etc.), it often becomes necessary to take immediate action to secure the processes and prevent potential additional equipment damage and avoid human or environmental impacts. This can require manual interaction with the process to open or close valves, stop pumps, and such. When these situations arise, they are extremely high stress events that can tax even the most highly trained employee. Such stress can result in human error (as human reliability is only estimated at 90% under normal conditions), delaying the securing or the system and increasing the potential for significant human and environmental impact. To be able to provide a person in the field with immediate feedback equipment locations, necessary actions, and required order of actions could be a critical step in maintaining both the actor’s safety and the safety of everyone nearby. As such, we propose utilizing object recognition and electronic visual feedback available from augmented reality to provide the person in those high stress environments simple, actionable feedback to decrease the chance of operator mistake while increasing the effective response rate of the person performing the task. This idea and technology could be expanded to provide feedback on contents of exposed pipes, tanker cars and liquid haulers to first responders more quickly and with direct information on how to handle the situation. By utilizing small databases of short, bulleted information, all the information could be stored in the local device. This would eliminate the need for internet connectivity, but could utilize connectivity, if available, to provide additional useful information. To this end, we have developed initial studies for evaluating the capabilities of AR for use when performing physical tasks under stressful conditions. This study compares memorizing tasks, hard-copy versions of tasks to be completed, and an AR-driven HUD display. Initial results indicate that an AR tool improves accuracy and speed while also freeing the users’ hands to complete tasks.

Initial results indicate that providing operators (student test subjects) with tools that eliminate the need for memorization, there is a significant improvement in task speed and completion accuracy. When these tools are utilized in a high-stress and/or low visibility environment, the impact of these tools increases significantly, thus justifying the concept and extension to real-world testing.