The virtualisation of the playout technology stack is a natural continuation of the technology evolution we’ve witnessed over the last decade. As an industry we’ve moved from using discrete single function appliances to multi-function appliances, through to software emulations of those appliances running on highly tuned, highly specific COTS hardware. Taking those softwares and running them on virtualised compute, storage and network infrastructure is the natural next step.
Each stage of this evolution has presented its own difficulties but in return, its own benefits. The move to ‘channels-in-a-box’ required vendors who once only needed to understand one aspect of playout to understand many. But as the technology matured, the benefits were realised just by swapping technology stacks out. The move away from large, multi-vendor deployments reduced the cost of purchase and deployment for an entire system. The ongoing cost of operation was also reduced as the physical space a channel occupied went down, along with its total power consumption.
As we make the move from ‘channels-in-a-box’ to ‘channels-in-a-cloud’ there is once again an expectation of an instant cost saving along with a whole host of other benefits. Agility and flexibility are often cited as enablers for business-transforming cost efficiencies and new monetization opportunities.
I’m not yet convinced that, for a channel that is constantly on air, ‘cloud’ brings any major cost benefit over its lifetime. However, this technological evolution is certainly a catalyst for some other benefits to be realised, although none of them can be gained through technical transformation alone.
To achieve the greater agility, increased speed to market and to open up the new monetization opportunities that ‘cloud’, in its many forms, promises, broadcasters also need to innovate operationally, and innovate commercially.
When we talk about virtualisation of technology we’re only really focusing on the tools. If all you do is change the material your tools are made of and nothing else, you cannot fundamentally change your business.
After speaking to many technology vendors at NAB this year who were exhibiting cloud based playout platforms, it was clear that whilst there was potential for a large change in the back end of any platform, the parts of the systems that people touch remain the same. If the people are the same and the processes that surround them are the same then changing what the tools are made of is not going to drastically reduce the cost of day-to-day operations.
Another benefit we hope to draw from virtualisation is a shorter time to market (i.e. get there quicker), and agility (i.e. turn it off if I no longer need it). These demands don’t marry up with the build time associated with a traditional playout deployment, the cost of which can usually be amortised over a long period. To get the benefit of this technology change we need to deliver it faster.
Without doubt some of the tasks that usually take place in a channel deployment like wiring, racking and stacking won’t be needed in a virtualised world. But configuration, testing, parallel running, and project governance are all still needed to deliver a reliable channel.
To achieve a shorter time to market and affordable agility we need to pre-can channel configurations; we need agile deployment teams with new governance processes; and those teams will need new tools for the configuration, launch and shut down of channels.
When we look at the role of the delivery teams and operational teams in linear playout, it is clear that virtualising the back end the tools isn’t enough to effect change, to deliver value. To realise the benefits of this new technology, you also need new processes.
The last aspect of our linear playout world that needs to change is the commercial landscape: how we buy from our technology partners, and how we sell our services to our customers. And I think this will be the most challenging thing to do.
Historically the technology that playout services ran on were chains of function specific appliances. In recent years as we have moved to using ‘channels in a box’ and the capital investment that used to be made in appliances has shifted to perpetual licences for somewhat less tangible software.
As we develop the ability to technically and operationally run channels for a short amount of time, or even for an unknown amount of time, the perpetual licence model no longer fits.
This is a huge challenge for technology vendors, particularly if the pay-as-you-go model starts to erode the sale of perpetual licences. But it has to happen and the successful vendors in this space may well be the ones who figure out the commercial model first, not necessarily those that get the technology right first time.
At Ericsson we are investing in research to solve the operational challenges of cloud based playout delivery and working with our partners to develop new commercial models so that we develop new services that drive genuine value for our customers. Whether it’s to protect their existing business or grow their business into new areas.
If you’re attending IBC in Amsterdam this year, please stop by and see us. You can find us in Hall 1, booth number D61.
Richard Cranefield, Strategic Product Manager, Playout