Lecture: Combining Computer Vision and Augmented Reality to Deliver Real Time Information and Troubleshooting

When a user needs any information related to hardware, whether configuring, verifying, or troubleshooting, they sift through documents, websites or forums. A couple of challenges associated with this process:

  •  Where/how to find the information
  •   Information lost in the clutter
  •   Time consuming process

These problems can be solved by using Augmented Reality (AR) to deliver real-time information. But today’s AR apps have a number of limitations:
      Use unique markers 

      On-the-fly markers

       Uses user-defined markers 

  We have solved these problems using Computer Vision (Deep Learning) to recognize hardware devices. Using Deep Learning, a bot can be trained to recognize hardware equipment. Once the bot correctly detects the object, you can use AR to deliver the required information. The solution is completely markerless. So far, at testing we have achieved 85% + accuracy on Cisco Routers. 
Furthermore, you can extend this functionality by implementing component level detection. Identify cables, ports, fans, diagnostic LEDs, and deliver pin-point and accurate information to the user.
This technology can be used in domains such as heavy industries, automobiles, oil & gas, and structural engineering. Just point your mobile at the device, tap on the screen, and see Augmented 3D animations of your device.