GE Software recently participated in the fifth annual Augmented World Expo (AWE 2014) in Santa Clara, California. Focused on augmented reality (AR), wearable computing, and the Internet of Things, this industry-leading event attracted innovators, futurists, scientists, creative agency professionals, investors, and industry watchers.
According to Wikipedia, AR is a live direct or indirect view of a physical, real-world environment whose elements are augmented (or supplemented) by computer-generated sensory input such as sound, video, graphics, or GPS data.
As part of GE’s investment in advancing the Industrial Internet, our Connected Experience Lab is exploring AR technology and its applicability for maintenance and assembly, training, and safety use cases. Imagine using an iPad to see an animated 3D model of a machine component (a fan), rendered over a real-time camera view of the actual piece of equipment (an engine). The view might looks like this:
Is Augmented Reality ready for GE?
Much of AR’s success thus far has been in the consumer markets – primarily in gaming. The GE Software team’s main objective for attending the conference was to gauge whether this technology is ready to support industrial use cases. Can AR work in harsh industrial environments? Can AR save our customers time and money? Will users embrace it? Does it support our vision for the Industrial Internet around asset and operations optimization?
Yes and no.
First, the yes: Delivering accurate visual guidance and reference to the exact placement and steps needed to overhaul a machine could save a field service technician time and allow her to do her job more efficiently and safely. A key use case for AR is to support repair, maintenance, and assembly of complicated equipment.
Now, the no: GE Software is committed to delivering information to users in context. What is the best form factor to present an AR experience in the field? Think of field service engineers who are often working on, and even inside of, equipment. These users really need to work hands-free. They won’t take on the burden of carrying around a device to use AR. This is particularly true for workers who MUST be hands free, such as surgeons in an operating room or repair technicians in small physical spaces.
One potential answer to this limitation is wearables, and specifically smart glasses. Some industry observers, however, suggest that interacting with more visual input through the lens or lenses of smart glasses can be very distracting.
Many of the leading smart glasses vendors were at AWE 2014 promoting hands-free AR with their devices. The GE Software team sampled some examples on demo but found the visuals to be fuzzy, choppy, and slow to render. It was clear that smart glass makers are committed to continuous improvements in visual quality, but for the moment, AR is best delivered via a tablet or smartphone.
Industrial Augmented Reality
This year’s event included several enterprise-related sessions. At “AR for Complex Manufacturing,” a panel of business representatives and academics spoke about how AR was enabling faster and safer production of equipment. They showed systems with AR views projected on glass between the user and assembly area and projected on the actual materials as well as viewed through iPads and other mobile devices. All participants said that the use of AR significantly speeds up work and helps transfer knowledge and that they had executive support to grow its use throughout their companies.
Don't touch my data
Like many of today’s burgeoning technologies such as cloud, big data, wearables, and the IoT, AR data possesses an interesting data privacy characteristic: the risk of engineering data being reverse-engineered from the graphics (2D and 3D). The good news is that much of the engineering data used must be simplified or geometrically reduced in order to render in real time; this requirement makes extracting data from a model very challenging. Other vendors are using 3D file formats that are next to impossible to reverse engineer. Without robust data protection and model simplification, the AR option will be very limited.