Media Cloud & Microservices

Most broadcasters today are using cloud services in some ways, but the cloud continues to change fast, and attractive new possibilities have opened up. The latest trend is towards serverless architecture and what is known as FaaS – function as a service – opening the door to an entirely new way of running media services.

MCMA provides a painless way of adopting serverless architecture and implementing media workflows in the cloud. Created as an open-source project, free to use and available on GitHub, it offers reusable services that can apply to all applications and infrastructures across multiple cloud providers.

2022

MAM based on MCMA (Q2 2022)
Demo at IBC (Q3 2022)

2021

Launch control: a web-based application to monitor and deploy complex cloud infrastructure (Q1 2021)
Cloud agnostic engine to develop workflows: Node-Red (Q2 2021)
Open-source libraries on GitHub for AWS, Azure, GCP (Q4 2021)
Standardisation at SMPTE (Q4 2021)

2020

Google platform development and integration in front-end (Q1 2020)
MCMA Developers Workshop @ Bloomberg (USA) (Q1 2020)
Code factorisation of FFMPEG and STT benchmarking (Q1 2020)
Presentation and demonstrations at EBU PTS 2019 (Q1 2020)
Workflows integrating transcribe services from the AWS, AZURE and Google Cloud platforms (Q3 2020)
Open-source libraries on GitHub for AWS, Azure, GCP (Q4 2021)

2019

Disseminate micro services-based technology among EBU members;
Provide tools and libraries to facilitate the development and deployment of workflows for media applications;
Support EBU members in deloping workflows using MCMA technology.

MCMA is a project of the EBU Metadata & AI group, under the umbrella of the EBU's Strategic Programme on Production.

REST APIs

Its main goal is to develop a set of simplified REST APIs with a minimum payload. These APIs allow integration of cloud and in-house based workflows, by combining microservices between (multiple) cloud platforms (AWS, Azure, Google) and in-house services.

Currently, the work is focussed on the Launch Control application; a user-friendly interface to help design and monitor microservices workflows.

Flexible workflows

So far the emphasis has been on Artificial Intelligence (AI) enabled services, such as speech-to-text, automatic subtitling, translation, text-to-speech, face recognition and identification, emotion detection, speech-to-text evaluation, FFMPEG media transformation. However, developers can easily duplicate and extend the code for other applications.

Multi-cloud

The MCMA cloud services provide an abstraction on top of provider-specific cloud services, providing cross-provider interoperability. Using the libraries, components and instructions available on GitHub, MCMA offers a turn-key solution ('implementation by example') for cloud infrastructure implementation, deployment and management.

MCMA has been successfully demonstrated at IBC and NAB events since 2017 and in a dedicated developer workshop in 2020.

Standardization

MCMA architecture allows to manage and process complex media workflows as a Job. The log management strategy is a critical challenge in a microservice architecture. A major challenge in building media solutions around multiple microservices is that status reporting is done in a wide variety of ways, making it very difficult to sort through these and determine the real status of the overall jobs being accomplished by these services. The MCMA team is working on standardizing Job Management with SMPTE:

Working group: https://kws.smpte.org/higherlogic/ws/projects/629/details.

Standard Proposal: https://github.com/SMPTE/st2126