Over the past few years, we at Crystal have been banging the drum for automation in many different parts of the broadcast chain.
By Alan Young
Traditional TV has never seen more challenges than it does today. Competition is rife from other traditional TV providers, as well as a plethora of over-the-top (OTT) offerings. In order to remain competitive, traditional TV providers are delivering more channels than ever before, and often supplement them with their own OTT offerings delivered to multiple platforms. This is causing a number of challenges when it comes to keeping track of content across all those channels.
Monitoring is already a critical part of all broadcast operations. In most broadcast centres, there will be a number of people with their eyes permanently glued to screens showing multiple sources of video, watching for anomalies. This is the way it has been done for many years because it has been the only way to determine that the right content is airing in the right place at the right time, and that it looks good. However, it has become more and more challenging to do this effectively as the number of channels, sources and rights restrictions have increased.
Over the past few years, we at Crystal have been banging the drum for automation in many different parts of the broadcast chain. Of course, it is becoming more commonplace in this industry and is having a massive effect in many parts of the workflow.
The more we automate, the more efficient, not to mention error-free the industry becomes, and the more broadcasters can focus on ensuring a great user experience for viewers. That said, we will never get rid of the need for human intervention nor should we want to. What we should be doing is automating the process of monitoring so that any change in the norm or expected outcome is logged and signalled to a human for verification and corrective action.
In the case of watching screens, Crystal believes there is a relatively easy mechanism that could automate this whole process. It relies on the fact that content in different parts of the broadcast chain is supposed to be identical some of the time. For example, the output of the main and backup video server should be identical all the time, and the return from the satellite or content delivery network (CDN) should be the same content outside of blackouts or ad avails.
With a person staring at content on multiple screens, it can be challenging to notice something that is not quite right. A computer can compare those multiple versions frame by frame continuously and very quickly pick out an anomaly — however small it may be — as long as the computer knows what the desired outcome should be. Differences that exceed a threshold could be very quickly brought to the attention of the operator, who can then determine the source of the problem and the appropriate action to take.
Automating monitoring in this way would have a number of very material benefits to a broadcaster or service provider. First, the accuracy of the monitoring is absolute, so any and all anomalies can be logged. This is very important for audit purposes.
Second, the number of monitoring points is very scalable, giving visibility along the entire broadcast chain through all distribution channels and to all screens.
Third, fewer humans (who can be located anywhere) can do a much better job of ensuring quality because they are only engaged when there is a problem, rather than having to be engaged “24/7” to look for problems that for the most part, are not there. Broadcasters get it right 99.99% of the time, so it makes sense to concentrate the manpower resources on the 0.01% of the time when things are not going so well!