By Bryan Kelly, Comcast Cable
As the landscape of “big data” resolves into tangible progress using machine-level (ML) data and artificial intelligence (AI) to proactively optimize the network, service providers using gateways, digital set-tops and related cloud components that are based on the Reference Design Kit (RDK) are reaping beneficial returns. This paper highlights multiple use cases involving RDK-derived data to provide context about, predict and fix problems before they impact customers, recovery with alacrity when problems occur, and apply anomaly detection to resolve “edge cases” and related challenges that occur across mixed cloud environments.
Operators around the globe are making use of machine-level (ML) data, derived from RDK devices, to feed artificial intelligence (AI) engines and algorithms. Those algorithms, in turn, fuel business- and care facing dashboards designed to continuously monitor and proactively address events that impact customers. That machine-level data is increasingly mined from in-home devices, including digital set-top boxes and broadband gateways, and based on the RDK’s open-source software stack inside system-on chip (SoC) silicon, both in set-tops and broadband gateways.
For the past few years, representatives from the RDK community have addressed Cable-Tec Expo audiences, detailing progress in business and technical operations and involving Artificial Intelligence (AI) algorithms fed by RDK-sourced machine data. This year, Comcast, as one of the RDK’s founders, will detail how it developed a customer experience optimization data platform, called Timeline, which builds a continuous and linear view of each interaction a customer has with the company – be it a series of nonscheduled equipment reboots indicating service problems, or an actual call.
This paper describes how the Timeline software platform, developed by Comcast, creates an empathetic and linear view of its customers’ experiences in terms of their interactions with the company including but not limited to signal quality, care-related incidents, product usage and assistance. It was built with APIs that are leveraged by many departments within the company, from product, to care, to technical operations.
Within, we describe three use cases, some in-production and some in-development. These illustrate how RDK-based data derived from set-tops and gateways is being used within the Timeline platform:
- Determinations of whether broadband gateways are healthy;
- Measurements of the aggregate and individual health of WiFi-connected devices; and
- Proactive analysis of error conditions within set-tops to obviate truck rolls and improve the customer experience.