Skip to main content
Discover the latest edge computing trends and technologies at ONE Summit, April 29-May 1 in San Jose | REGISTER

Applications at the Industrial Edge

By February 17, 2020July 21st, 2020Postcards from the Edge

By Vineet Anshuman
Co-founder, Cloudlyte

& Vikram Balimidi
Director Of Product and Marketing, SD-WAN Services


Major strides have been made in the last few years in making Augmented Reality (AR), Virtual Reality (VR) and Mixed Reality (MR) more accessible and affordable, enabling the market to move towards mass adoption. What’s also becoming apparent is that enterprises and industries can really benefit from these technologies (for a primer on the differences between these technologies, click here).

According to IDC, the AR/VR market spend was around $17.8 billion in 2018 and growing at over 95%. Funding for these technologies is only increasing YoY, and the use-cases are moving beyond gaming, media and entertainment. Industrial use-cases are fast emerging as key drivers for AR/VR technology, encompassing uses ranging from design, engineering, training, and field services to live remote support.

Among these related technologies, AR is expected to be implemented more widely, with over 66% of respondents in recent Capgemini research agreeing the technology will be more impactful in real-life scenarios.

Specific use-cases are emerging in manufacturing, heavy industries, retail, utilities, automotive and construction sectors. AR is  extremely useful because it enables the digitization and automation of workflows, enables companies to train technicians on technology and safety, and ultimately resulting in reduced manufacturing and operational errors, improved operational efficiency, enhanced collaboration and faster innovation. The end results are significant cost savings and enhanced top line results.

  • Boeing has used augmented reality for technicians, increasing productivity by 40% and reducing wiring production time by 25%.
  • Walmart has used VR technology  to train employees; they’ve acquired 17,000 Oculus headsets for training in over 4,700 US stores. 
  • Other use-cases around remote-assistance and product training have shown improved efficiency by reducing the time for support calls by 25% and an attendant cost reduction of 35% by cutting travel expenses, as well as gains in product quality. 

AR is definitely a growing space and the ecosystem to support these use-cases is emerging quickly. Collaboration amongst the purveyors of hardware, software, services and infrastructure ensure the best user experience is always being delivered. 

However, barriers to AR/VR still exist.oday, AR/MR experiences are predominantly delivered from a centralized cloud hundreds or thousands of miles away from end users. The issue with hosting these applications in a centralized cloud is that as the use-cases move towards more real-time requirements and low-latency (within 10ms-20ms), moving the cloud closer to the end-user is an absolute necessity.

Rendering and processing complex hi-definition 3D models and visualizations in a centralized cloud and continuously delivering the results to user headsets limits the kinds of immersive experiences that can be delivered. Not overcoming this barrier will create a significant hindrance to user adoption, and ultimately the successful implementation of enterprise AR use-cases.

Today’s enterprise AR/MR applications and services require a highly responsive compute infrastructure and 24/7 availability. Some examples from live industrial environments show some of the critical challenges:

  • An airline engineer who is less experienced needs to interact with a more qualified engineer to troubleshoot and repair an engine issue. An AR application enables the technicians to interact and delivers advanced visualizations that can help reduce errors, reduce costs and speed up the repair process. When this scenario was tested, the researchers found the delays in the real-time interaction and visualization led to  user acceptance issues, which impacted adoption.
  • Many use cases require the projection of large and complex 3D industrial models, combined with real-time video assistance, all overlayed onto an image of the physical environment around a worker. In order to deliver a good experience that won’t disrupt the viewer’s balance, real time calculations must be made to incorporate the user’s FOV (field of view), motion and orientation.

To deliver a low-latency experience, services and instances of the applications need to be deployed near the end-users where it is most relevant — near the edge of the network. 

Allowing AR applications to be built and delivered across the cloud-to-edge continuum requires tools that help developers  access and consume the best resources in the ideal locations. 

Next-generation edge computing infrastructure should move beyond the current deployment models to deliver a PaaS-like experience characterized by:

  • Support for standard languages 
  • Act as an extension of cloud tools and development methods
  • Deploy multiple workloads including containers and serverless
  • An ability to support stateful applications
  • Are fully programmable by the developer
  • Intelligent infrastructure and application deployment
  • Fully secure and encrypted 
  • Provide monitoring, troubleshooting and diagnostics.

The edge PaaS to deliver this should be flexible and scale as needed, with the ability to process large portions of an application when needed.

AR application developers and platform providers need to be offered a consumable model to access the infrastructure whenever needed with minimal effort and integrating into their current CI/CD development pipeline.

For AR application developers and platform providers, edge computing offers a critical component to solve the use experience issues impacting AR applications and adoption and should be as easy to program as any other cloud application.