Industry Insights: How AI is impacting broadcast production workflows

By Jacob Billingsley

Subscribe to NewscastStudio for the latest news, project case studies and product announcements in broadcast technology, creative design and engineering delivered to your inbox.

We delve into the transformative role of automation, machine learning and artificial intelligence within the broadcast industry in the second installment of our Industry Insights roundtable. This discussion uncovers the real-world applications of these technologies, highlighting how they are reshaping everything from content production to viewer engagement.

Participants explore the efficiencies AI introduces in automating routine tasks, the advancements in content personalization, the pivotal role of AI in predictive analytics and its impact on live broadcast production and multi-platform content distribution.


Key takeaways from the Industry Insights roundtable

  • AI, machine learning, and computer vision are key in automating tasks within broadcast production, significantly saving time and enhancing efficiency in processes such as app testing and video quality measurement.
  • Through automation and AI, the quality of broadcast content is improved by automating tagging for clip selection and utilizing AI upscaling to transform standard definition content into higher resolutions.
  • Real-time analytics and decision-making in broadcasting are being modernized with AI tools that analyze viewer engagement and preferences, enabling broadcasters to deliver content more effectively and optimize ad placements.
  • AI plays a significant role in predictive analytics for content performance, using KPIs from video quality analysis to train algorithms for detecting and anticipating service anomalies.
  • AI technologies are streamlining content management and archiving by automating metadata tagging and enhancing search capabilities, which improves monetization opportunities and access to relevant content.

How is AI being used to automate routine tasks in broadcast production?

Mathieu Planche, CEO, Witbe: AI, machine learning, and computer vision algorithms are all being used to automate practical tasks and save teams valuable time. This has especially taken off in the processes of app testing — where AI allows teams to script test scenarios ten times faster — and video quality measurement, where tests now assess video quality with the same criteria as a human viewer. Moving forward, automation will continue to be an area where AI can make practical differences in substantial ways.

Abhijit Dey, general manager and COO, Perifery: AI automates routine broadcast production tasks, such as transcriptions, translations, summaries, and keywords, which were previously time consuming. AI also simplifies object and facial recognition, enhancing content searchability and efficiency.

Paddy Taylor, head of broadcast, MRMC: I think we are only scratching the surface of what’s possible. We have seen significant advances in reducing the time it takes to create automation events. With our robotics for camera automation the use of Polymotion Chat (our CV tracking system) uses AI and prediction to make the camera movement appear human (typically these systems can ironically appear robotic) which allows the operator more freedom and the ability to manage more cameras.

Lara Guerard, principal solutions director, TMT Insights: AI is being used in broadcast production to automate and speed up tasks such as video editing, where it can quickly sort and edit footage based on pre-set criteria or perform video/image fill, or prompt-based creatives. It’s also used for automatic transcription and translation of dialogues, significantly speeding up the subtitling and localization process, making global content accessible for a wider regional reach. These AI-supported applications streamline the production process, reduce manual effort, and increase content efficiency.

Vinayak Shrivastav, co-founder and CEO, VideoVerse: Previously, production teams had to manually resize, brand, and distribute broadcast content to digital channels like social media. AI-based video editors automate all of these routine tasks through features such as auto-resizing, auto-personalization, and auto-distribution. By saving time on these tedious tasks, production teams are free to spend more creative time and energy on more important production workflows that directly drive viewership and audience engagement.

Advertisement

Jay Prasad, CEO, Relo Metrics: M&E organizations are increasingly realizing the benefits AI and machine learning (ML) technology to streamline workflows and enhance efficiency in a range of business process functions, from long-term market forecasting to daily production and operations management. Manual tasks that may have previously taken days or weeks, even months, to complete can now be done in minutes, relieving staff from the burden of tedious and time-consuming functions and freeing them to focus on “higher value” activities that more directly contribute to business success or customer satisfaction. Here’s a perfect example: With its ability to analyze video and audio content and automatically generate descriptive metadata, AI has proven its effectiveness in enhancing content and data searches, accurately detecting, and classifying people, objects and or locations within a piece of content. 

In what ways are automation and AI being used to improve content quality in broadcasting?

Gerhard Lang, CTO, Vizrt: AI-driven automation reduces the need to have people spend too long on tasks that are time-consuming to produce a high-quality broadcast show. It’s about improving the usability of the systems needed to create shows and content, which invariably eliminates boring tasks and reduces potential error sources through automation. For example, voice-to-text commands to drive the gallery’s vision mixer, graphics, and other behaviors — and these allow the adjustment of content quicker and with a lower risk for error, which improves the overall quality of the broadcast.

Steph Lone, global leader of solutions architecture for media and entertainment, Amazon Web ServicesVideo encoding/decoding has grown smarter in recent years. Take AWS Elemental Media Live’s quality-defined variable bitrate (QVBR) control for instance. It leverages ML to vary the number of bits used to produce a compressed video to meet the perceptual quality the user wants to achieve, ensuring the highest quality live stream at the lowest cost.

Mathieu Planche, CEO, Witbe: The process of improving content quality is rife with repetitive and labor-intensive tasks. Teams must first measure it, then address any issues they find, and finally measure it again to verify their efforts. With automation and AI technologies, the time spent setting up measurements can be drastically reduced, freeing up teams to dedicate more time to the actual engineering work of improving quality.

Abhijit Dey, general manager and COO, Perifery: Automation and AI can dramatically improve the quality of broadcast content with automated tagging for quicker clip selection and AI upscaling. AI tagging utilizes facial and object recognition to enable efficient sorting, while AI upscaling transforms standard definition content into HD or 4K, maintaining original fidelity.

James Fraser, VP for U.S. sales, Newsbridge: A well-trained AI model ensures the consistency of media logging will be at a much higher standard than that of a human media logger. The reason is simple: AI performance isn’t dependent on how many hours a person has worked, how much sleep they’ve had, or how much caffeine they need to perform their best. The benefit of AI automation for human media loggers is that it removes the heavy lifting and frees them up to perform higher value, more creative tasks such as curating media collections, producing highlight reels and improving their organization’s AI model.

How are broadcasters leveraging AI for real-time analytics and decision-making?

Lara Guerard, principal solutions director, TMT Insights: In a non-linear and digital environment, AI is being leveraged for real-time analytics and decision-making by utilizing AI tools that analyze viewer engagement and preferences. With access to 1st-party data, real-time data-driven decisioning determines which content to deliver to the viewer at the right time while also assisting in the dynamic optimization of ad placements and scheduling, therefore enhancing revenue streams. This real-time insight into viewer behavior and market trends enables broadcasters to make informed, immediate decisions to improve the viewer experience and operational efficiency.

Jay Prasad, CEO, Relo Metrics: Speaking from our perspective in the sports marketing segment, we’re seeing a continued shift from traditional broadcast viewership to more streaming, enhanced digital content viewing experiences and a rise in programmatic advertising. Our industry has been forced to modernize its measurement methodologies as well, focusing on the convergence of technology, data-driven software, and media assets to create a self-sustainable ecosystem that can lay the foundation for the future of sports currency. To truly optimize and maximize creative assets in sports and entertainment, we have to go far beyond just traditional valuation metrics that have largely been done in silos.

What role does machine learning play in predictive analytics for content performance in broadcasting?

Mathieu Planche, CEO, Witbe: Something we’ve discussed when looking at the future of AI is how the KPIs that we already measure for video quality analysis could be used to train a machine learning algorithm to detect and anticipate anomalies on video services. This represents an area that we believe would greatly benefit from machine learning and AI in the future.

What advancements in AI are driving the personalization of broadcast content?

Lara Guerard, principal solutions director, TMT Insights: Content broadcasters are prioritizing personalized viewer experiences to capture and retain viewer attention. AI advancements such as machine learning algorithms for analyzing viewer preferences, natural language processing for understanding viewer interactions, and predictive analytics for forecasting content trends are some examples of the advancements in AI technology driving this personalization. For example, AI-powered recommendation engines tailor content suggestions based on individual viewing habits, while computer vision technology enhances content discovery by analyzing visual elements within videos.

Vinayak Shrivastav, co-founder and CEO, VideoVerse: Data Sports Group has estimated that almost 80% of viewers use a second screen while watching live broadcast content. For example, a fan watching a football game may also be following the box score online. AI can make this secondary screen much more personalized: Because broadcasters can auto-generate clips easily, they can publish a greater library of content to cater to every person’s individual viewing preferences and needs. 

Jay Prasad, CEO, Relo Metrics: AI algorithms simply give brands more powerful storytelling tools, contributing to the analysis of viewer behavior patterns to help predict content preferences, for example. Also, broadcasters and content producers can tailor programming and advertisements for maximum audience engagement.

How is AI being used for more efficient content management and archiving in broadcasting?

Steph Lone, global leader of solutions architecture for media and entertainment, Amazon Web Services: Many media and entertainment customers are using AI and ML for automated metadata tagging and enhancing available metadata. For instance, Amazon Rekognition can detect objects and people in video content, and it can also find text and perform “video segmentation,” an analysis of the video structure that creates a timeline of significant events (i.e. color bars, black frames, title credits, introductory material, logos, channel identifiers and more). That data can be stored alongside the video itself, and used to find segments of the video that feature people or specific topics of interest.

Advertisement

Abhijit Dey, general manager and COO, Perifery: AI revolutionizes broadcast content management and archiving by enabling close integration with content. The latest AI tech enriches newly added content with metadata, optimizes storage across tiers, and provides editors with extensive information about data without backup or space concerns.

Lara Guerard, principal solutions director, TMT Insights: When broadcasters incorporate AI into their content management and archival workflow, it generates metadata tagging through intelligent services such as speech-to-text, facial recognition, etc., and enables advanced search capabilities, providing easier access to relevant content. This increases monetization opportunities and creates a faster time to market.

James Fraser, VP for U.S. sales, Newsbridge: With AI, it’s possible to process and index massive volumes of content at speeds that are simply unachievable by humans, at a much lower cost than manual digitization and logging methods. AI green lights large archive preservation and monetization projects that were previously deemed untenable, allowing organizations to complete these tasks in considerably shorter time frames.

Charles Dautremont, CTO, Cinedeck: Perhaps one of the most controversial features AI has is the ability to create automated content, which can then be managed to create a manufactured content workflow. The automation of creative content offers a great deal of efficiency and benefits particularly in post-production and live-production workflows, providing the integration is handled correctly. AI is capable of archiving content, offering further opportunities to media asset management (MAM) systems by tackling manual tasks, captioning/editing content and aiding in the build of an entire broadcast workflow that flows and functions with the benefits of AI integrated within it.

How are AI technologies helping in managing multi-platform content distribution in broadcasting?

Abhijit Dey, general manager and COO, Perifery: AI simplifies multi-platform content distribution by automating content packaging and delivery for video streaming platforms like Amazon, Netflix, and YouTube. Broadcasters are embracing AI for its ability to ensure compliance with platform specifications and broadening distribution reach without increasing manpower.

Lara Guerard, principal solutions director, TMT Insights: For multi-platform content distribution in broadcasting, AI is able to optimize content versions for different platform deliveries through tasks such as automatic formatting and resizing. They employ predictive analytics to determine the best times and platforms for content release, maximizing audience reach while also personalizing content for diverse audiences across platforms, enhancing viewer engagement. It can streamline workflows by managing and synchronizing content across various channels, ensuring a consistent and efficient distribution process.

What role does AI play in enhancing live broadcast production?

Gerhard Lang, CTO, VizrtAI can bring an idea into an executable task. It also provides helpful information about a specific topic, and as we’ve seen frequently, AI can suggest setups based on previous feedback — this helps people make decisions based on what is visually more appealing, entertaining, or easier for the audience to understand. The nature of AI is that it is constantly learning and therefore the quality of a show and content creation will improve over time.

Vinayak Shrivastav, co-founder and CEO, VideoVerse: Some broadcasters may assume that AI is only relevant to archival footage. This may have been true a generation ago, but AI can now ingest live streams and apply these same technologies on the fly. This enhances live broadcast production because viewers can now enjoy content that has seemingly gone through extensive post-production processes all in near-real time.

James Fraser, VP for U.S. sales, Newsbridge: AI technology empowers vendors to automate a wide range of manual processes, including the indexing of live streams. AI that automatically detects and indexes faces, objects, actions, locations, and context saves video editors significant time, enabling them to perform a simple text search and pull up the exact clips they need to build a story and publish it, capitalizing on the moment.

How is AI assisting in audience engagement and interaction during live broadcasts?

Steph Lone, global leader of solutions architecture for media and entertainment, Amazon Web Services: We are seeing many companies utilize the metadata they extract both from their live and archival content to curate personalized consumer experiences, driving higher viewership duration and ad delivery. Second screen experiences are about to get a huge boost from AI. If you join a sport broadcast late, naturally, you want to be able to catch up as quickly as possible. Machine learning technology can now generate a highlight reel of the broadcast specific to the time you missed.

Vinayak Shrivastav, co-founder and CEO, VideoVerse: When viewers watch an amazing moment, such as a dunk in a basketball game, they naturally want to see it again. Unfortunately, because broadcasters have limited resources, they cannot produce all the highlights that viewers would possibly want to see during a given game. With an AI-based video editor, broadcasters can produce exponentially more highlights on a real-time basis, driving engagement with fans eager to rewatch and interact with them.

Jay Prasad, CEO, Relo Metrics: Live broadcast coverage has turned into a competitive race to continually improve and keep giving viewers that “you are there” feeling. With more automated graphics and increasing applications of AR, VR and MR, AI can generate graphics and visual effects in real-time to enhance the overall visual appeal of broadcasts, playing a key role in the ongoing battle to attract and retain eyeballs across all platforms and audience demographics.

Explore more from this roundtable

Subscribe to NewscastStudio for the latest news, project case studies and product announcements in broadcast technology, creative design and engineering delivered to your inbox.