Verification of Electrical Grounds/Bonds Using Computer Vision (2020)

By Shawn Kercher & Jacob Hallberg, Comcast Innovation Labs

Electrical grounding and bonding, as a means to provide a low-impedance electrical connection between two or more metallic bodies that are normally not current-carrying, has long been a necessary task for the industry. Electrical grounding prevents circuit overloads and removes dangerous ground-fault voltage on conductive parts. Primarily, grounding and bonding is a mechanism to protect homes and businesses, and their contents (including people), from electricity surges, such as from lightning.

When all of the communication and power grounds are working at a home or business, and a real “good grounds,” any stray current in the system goes to ground and no voltage differences exist between them which is the desired state. However, when grounds are bad, the current can and will find its way into the CATV network. When voltage appears in the CATV network the voltage potential problem appears to “drop the voltage” of anything connected to the network.

Set-top boxes, modems, gateways, or, in the case of large MDUs, line extenders all have seen the drop in voltage. Stories abound from technicians taking the hit of 90-volt shocks from power passing taps, or 120-volt shocks when hooking cable up to taps in MDU lockboxes – which makes for a decidedly unpleasant site visit! Proper electrical grounding and bonding is as equally or more important to business/enterprise customers especially when lightning hits a facility, knocks out half of its servers, and turns out to be a grounding issue – as in, “our fault” (pun intended.) This paper describes the use of Computer Vision (CV), Artificial Intelligence (AI) and Machine Learning (ML) to provide technicians with a visual verification that gives them a high level of confidence that the structure (the home or business) they’re visiting is properly grounded and bonded, using an app developed at Comcast Labs, on their work-issued iPhone or iPad. The paper covers the background and history of grounding and bonding; why those activities continue to be vitally important in communications network performance; and the importance of images and image annotation to build a production-grade database of proper grounds/bonds that can be put to work by field technicians when assessing a home or business.

By clicking the "Download Paper" button, you are agreeing to our terms and conditions.

Similar Papers

Enforcing Social Distancing Using Computer Vision and Deep Learning
By Wael Guibene & Hossam Hmimy, Charter Communications
Improving Pedestrian Safety using Computer Vision, Machine Learning and Data Analytics
By Parmjit Dhillon, Mohamed Daoud & Hossam Hmimy, Charter Communications Inc.
Augmented Reality for Network Visualization
By Joshua Seiden & Nishesh Shukla, Comcast Innovation Labs
Photon Avatars in the Comcast Cosmos: An End-to-End View of Comcast Core, Metro and Access Networks
By Venk Mutalik, Steve Ruppa, Fred Bartholf, Bob Gaydos, Steve Surdam, Amarildo Vieira, Dan Rice; Comcast
Key Learnings from Comcast’s Use of Open Source Software in the Access Network
By Louis Donofrio & Qin Zang, Comcast Cable; Vignesh Ramamurthy, Infosys Consulting
Two Years Of Deploying ITV/EBIF Applications – Comcast’s Lessons Learned
By Robert Dandrea, Ph.D., Comcast Cable
Intel’s Vision Of Sports Immersion
By Guy Blair and Rajeeb Hazra, Intel Corporation - Intel Architecture Labs
Cloud-based Dynamic Executable Verification
By Rafie Shamsaasef, Aaron Anderson & Sasha Medvinsky, CommScope
Computer-Aided Analysis Of Coaxial Cable Attenuation As A Function Of Frequency And Temperature
By Jacob Shekel and Eric Winston, Jerrold Electronics Corporation
Simplifying Field Operations Using Machine Learning
By Sanjay Dorairaj, Bernard Burg & Nicholas Pinckernell, Comcast Corporation; Chris Bastian, SCTE
More Results >>