The question is who will buy them. The IT giants could be forgiven for getting cold feet: they still do not seem to have any sure navigational skills through the media jungle. Only Amazon’s acquisition of Elemental seems to have succeeded by dint of impeccable logic.
Back at NAB, while many of the big IT names were there, they were in general keeping a fairly low profile, often promoting the works of partners. Which was a shame, frankly, because a couple of the hot topics were around technology issues.
I was on a bit of a mission to find the state of the art in cyber security for our business. It was not easy. There was a fair bit of shifting in seats and changing of subjects whenever I asked the question.
I was not alone in finding some frustration. Andy Wilson of DPP told me that trusted partnerships were critical, and that includes the technology platforms that broadcast and media companies choose. “Vendors have to demonstrate their commitment to protecting broadcasters’ security,” he told me.
Increasingly, media companies recognise that content needs to be protected in transit, at rest (while stored on a server, wherever that server is and whoever owns it) and at work – while being processed. Indeed, a new company to me, Fortium, was showing an encryption platform which includes a full audit trail so you know at all times who has touched something and what they have done to it.
“There are different levels of security: around access, around transmission and around the integrity of the data,” Francisco Ontoso-Ramos, CTO of Object Matrix told me. His technology provides encryption at rest using secure keys, so you can store media on non-trusted servers. This could save cost, which would allow you to store multiple copies of the data in different locations, making the content store self-healing.
A new technique which is potentially at great risk of cyber attack is remote production: carrying all the cameras and microphones back from a remote location to a control suite at the broadcaster’s headquarters. One of the most developed remote production networks was implemented by Telstra for NEP in Australia, linking all the major soccer, rugby and Aussie Rules stadiums in the country to production centres in Sydney and Melbourne.
The links use technology from Aperi. I asked company chair Eugene Keane how they could possibly secure such a network. He pointed to the discrete, FPGA-powered hardware they use, and described how you can configure it to have trusted and untrusted sides on the same board.
“We can build 100% perfect firewalls using our NAT configuration app,” he said. “Then we configure what we allow in. We might define the system to only allow 2022-6 streams through, with packet inspection on the fly.”
Last word (for now) on security goes to Charles Shiflett of Aspera, who told me “Our goal is to make it impractical to break into the content – even if you are the NSA.”
The other IT buzzword that broke into the media big league at NAB this year was artificial intelligence. Even though most people were not talking about artificial intelligence, or at least a not very intelligent version.
Machine learning which can identify images was a very common suggestion. Chris Witmayer of Nascar, the motorsports people said “Although we have an entire archive that goes back to the 1930s, we can’t actually find anything efficiently. If you can’t find anything you can’t sell it and you can’t make money. So this is big for us.”
AWS Elemental is adding machine learning to its cloud encoding for the royal wedding. Sky will use it to scan pictures of guests arriving at the chapel, identify them, then link to more information. Viewers of the online and mobile services will be able to click on a guest to “enrich the live video content with facts about the attendees”.
Tedial has developed a neat system with a European broadcaster which takes machine learning one stage further. It analyses a football match in real time and automatically creates clips, which can be offered to the director and posted to social media.
The clever bit is that it also analyses the response on social media. If a clip goes viral, then the machine learns to make more like that. Lots of down-votes and it knows that idea bombed.
I also talked to Qligent which has recently run a big data project with a major telco in the States, aiming to find the silent sufferers, those who are not grumpy enough about service failures to complain, but could be grumpy enough to change suppliers at the next contract.
The project analysed several terabytes of data a day, drawn from Qligent’s data measurement probes, to see where there might have been less than perfect online delivery. The operator can then proactively contact the subscribers to say “we’re aware there might have been some technical issues, and we are working to see it does not happen again”.
Andy Quested of the BBC told me that he sees this sort of application as the real benefit of artificial intelligence. He feels that it is the route to a real measurement of quality of experience. I certainly hope so.
Finally, and talking of quality of experience, there was no doubt that this was a very quiet NAB indeed. There is still no announcement on the NAB press site, but the official figures for registration – they go through a statistical validation to get to an estimate of attendees – had a 10% drop this year over 2017.
One detail not immediately obvious from the raw numbers is that the number of exhibitors rose from 1518 last year to 1718 this year, or about 13 % more. Given that those 200 exhibitors are likely to be small this might be only around 1000 people, but that makes the drop in visitor numbers that much greater.
Does this mean the industry is falling out of love with exhibitions? Or does it mean that the industry is in a period of consolidation while we work through the challenges of moving to IP connectivity and software-defined architectures?
Maybe we will know more at IBC in the autumn.
|