Summit AR: Development of Augmented Reality Tools for Data Monitoring

Summary

During the summer of 2018, I interned at Oak Ridge National Lab (ORNL) in the Advanced Data and Workflow group. The lab had become interested in augmented reality (AR) and its potential applications. I was tasked with developing an AR application involving the new super computer Summit. Summit is the world's fastest computer and was just coming online while I was at the lab.

A lab mate and I developed a Hololens application that rendered the virtual cabinets on top of the physical machine. We did this to augment the physical space with potentially helpful data. We did not yet have access to real data so we mapped fake node temperature data onto the  nodes. A user would be able to walk through the machine room and see hot spots (even if the source is in another aisle).

Check out this article written about the project!

Background

Augmented reality is an emerging technology that allows us to augment the physical world with virtual information. This augmentation can be achieved through the use of transparent displays or by overlaying graphics onto a screen displaying a camera feed. The combination of physical and virtual experiences places AR on a spectrum commonly known as mixed reality. Mixed reality ranges from entirely virtual experiences to entirely physical. The mix of physical and virtual offers several advantages over traditional information displays. AR systems allow users to maintain focus on the physical environment while critical information is displayed in their peripheral. Virtual objects or holograms can also be a focal point in AR, allowing users to see through, move around, and inspect objects in ways that are not possible on 2D displas. For these reasons, AR has already seen use in education, entertainment, data visualization, maintenance, order picking, military applications, and in the medical field.

Researchers depend on data to produce scientific results and innovations. Data is an essential part of research workflows and tools to gain insight from data are needed. These tools should allow users to analyze, inspect, and interpret data quickly and effectively. The source of and use for data varies from project to project. In some large professional lab environments, real-time data from machines such as supercomputers is critically important. There is also value in lab equipment and machinery information to enhance the experience of visitors to the lab.

Below is a scaled down version of Summit, rendered in the Hololens 1. Users are able to click on each cabinet to take a closer look. The colors (light blue to dark blue) represent the node temperatures, but could also represent any other metric one might want to know about Summit, like computation load or power consumption.