BVE 2018 Round-up

By Alex Snell

The ExCel Centre London has just hosted the 2018 Broadcast Video Expo, an event that soldiered on despite the inclement weather, and with huge importance for the UK media and entertainment industry. Over 300 vendors and organisations showcased their latest technology, and business leaders provided their unique insight on technology and business trends. A number of key themes were very prevalent at BVE 2018, and they will be explored through this blog.

Firstly, how is new technology adopted?

Before getting into the trends of BVE, it is important to consider how markets respond to technology announcements. This both allows a judgement to be made about where a technology is in its adoption cycle, and more importantly, to cut through the buzz words to gain an idea of how developed and accepted a technology is.

Humans do not like change, and whenever new technology is introduced, the same thing happens in most cases. Most consumers start as sceptics, but through trial and error, the advances made by a small portion, and the achievements that become visible, most eventually become advocates. Sociologist Everett Rogers’ 1962 book “Diffusion of Innovations” is a relevant and applicable way of analysing the state of the industry today. The theories laid out have been proven time and time again, and the ideas are as relevant today as they were in 1962. There is a demonstrable correlation between new technology awareness, and new technology adoption. Figure 1, an adapted illustration of a Rogers’ diagram, illustrates the relationship between technology awareness and technology adoption rate (his example is based on farmers adopting weed killer, but the concept is what is important).

Figure 1 – Adapted from Rogers’  “Rate of awareness-knowledge, rate of adoption, and length of innovation-decision period” figure 5-5, page 204.


Events such as BVE help to increase awareness, and set the heading for industry trends. This makes BVE an important catalyst for change.

Another useful and widely accepted tool for market analysis is the “Law of Diffusion of Innovation”[1]. This a perfect illustration of the process that happens every time new technology is introduced, and live IP, a hot trend at the moment, is certainly no exception. Figure 2 is the best way of illustrating the law, which represents market share as area under the curve. Some notable innovators and adopters have been listed, in this case using Live IP as the new technology. Live IP is considered in the following explanations of each of the stages of adoption.

Figure 2 – Adapted from Rogers’  “Adopter categorization on the basis of innovativeness” Figure 7-2, page 247


There are 5 key parts of the law, with the first being the innovators.


2017 was the year of IP innovation, based on the developments from the VSF on the TR standards of 2015. Companies began investing in 2110 product development in in 2016/17, with interop’s as early as April 2017 between vendors to prove that their interpretations of the pre-release draft 2110 specification were good. These innovators were able to provide evidence of their systems working in places such as the Timeline TV OB truck. A main feature of BVE 2018, UHD-2 has 2110 technology at its core. The innovators represent 2.5% of the market share.

Early Adoption

2018 will be the year of early adopters. While Timeline and Arena were examples of very early adopters, more infrastructures will be upgraded to 2110 capability. The early adopters represent 13.5% of the market share. The establishment of a SMPTE specification for Live IP will be another important milestone in prompting early adoption, to ensure interoperability in outside broadcasting. This may be comprised of SMPTE 2110, AWMA IS-04/05/06, and deal with unanswered questions like how to clock two OB trucks to the same PTP Grandmaster, or what sender model should be used (wide or narrow?).

The Chasm

There is a tipping point at approximately 16-20% of market share where the market accepts a technology called the chasm. This breakthrough is considered the point of no return, where with continual development and adoption, a new trend will be accepted by the majority of the market.

Early Majority

Assuming that the continuous development of live IP continues through 2018, and the critical mass of beyond 20% is surpassed, 2021-2024 will likely be the time span in which an early majority is formed, at least in the UK. The BBC Wales studio complex will be a major landmark in the UK for 2110 and live IP.  As explained previously in figure 1, this tends to have the biggest delta between awareness and adoption (space between the lines on the graph), as the hesitant majority observe and learn from the early adopters before deciding when to upgrade. This segment represents 34% of the market share.

Late Majority

It is much harder to predict when a late majority will be formed than an early majority, and the biggest barrier to this establishment will be financial. We will not see the late majority until the vast majority can afford to replace their SDI infrastructure, or until lack of SDI development and interoperability means that its impossible not to have an IP system.


This final 16% of the market share will likely exist on the fringes of the industry where money is so tight that it is not possible to replace SDI systems until they become obsolete, or break and need replacing, once IP is the only option.

BVE’s biggest trends

Artificial Intelligence, beyond a buzzword?

There is one buzz phrase that was mentioned again and again at BVE – Artificial Intelligence. At the macro-level, AI was by far the most “anticipated” topic (just ahead of “the cloud”). The progression of AI within media has been extremely rapid in the last 3 years, to move it from a proof of concept in complex sandbox environments, to a valuable service in a broadcast workflow. Although it is already deployed for many use cases in areas such as medicine and finance, AI is now truly breaking into the media space, and the power it brings will be truly transformative.

When considering making workflows more intelligent, there are three main pathways, which can easily be blurred into the wider “AI” lexicon. They are automation, machine learning, and true artificial intelligence. Workflow automation has been adopted for a number of years, and is simply setting rules for processes. It is easy to get this and “AI” confused. Automation is pre-defined, and not dynamic. While efficient and impressive, there is nothing “intelligent” about simple automation.

Machine Learning (ML) is the key facet of AI that is usually meant when talking about AI in media workflows. This process relies on training a model to pattern match or characterise samples that it is given. This is what broadcasters are currently implementing. There are an abundance of use cases for ML in media workflows, from converting speech to text, to automatically tagging content with metadata.

The third is “true” AI, which is nowhere near where scientists think AI might take us – to have like-for-like human intelligence. Machine Learning is very much still a brute force method of replicating intelligence. True AI will not be. We are many years, if not decades away from this level of complexity.

In terms of adoption, AI is still in its very early stages, and is still in the innovation stage. That said, there were cases of AI in many scenarios. Endemol Shine stated that they used AI technology to collate, analyse, and produce metadata based on mass data sets from housemates in the Big Brother series. Viacom stated that they are using ML to translate speech to text. IBM’s representative said IBM’s Watson was used to cut highlights in the US open tennis, and to analyse a basketball team’s performance.  A reassuring and recurring notion was that AI would be used as a tool, and not a replacement, for human effort. How closely this policy is followed may be up for interpretation, particularly in non-creative roles.

However, things are moving very fast, and by IBC 2018, it may have already crossed the chasm and reached critical mass. In the media space, AI looks to be a revolution, not an evolution.

Every silver lining has a cloud

As expected, cloud technology was heavily represented, with speakers on almost every stage making mentions of cloud technology. The last 18 months has seen many major advances in cloud technology, which is enabling this trend. In 2016 it was very much an immediate future, whereas now it is a present reality. Compared to IBC 2016, the use of the cloud within the Media and Entertainment industry has also shifted focus. This movement has been from using cloud for storage and archive, to now also using the cloud to “spin up” applications and micro-services. Although the adoption of the cloud for storage continues, the future will be full of leveraging the cloud for everyday business processes such as transcoding, running user applications, or even playout. Additionally, more and more lower-tier players are providing offerings that challenge the main players such as AWS, Google Cloud Platform, or Microsoft Azure. These providers fill in the gaps that the main providers do not address, or build compelling service offerings on top of the main providers.

Cloud Cover

Another large and unexpected change compared to IBC 2016 was the attitude to cloud security, which has been seen as the biggest barrier to cloud adoption for a long time. In its infancy, cloud adoption was held back by nervous business leaders, who wouldn’t dare allow their sensitive and costly assets to be hosted in someone else’s data centre. However, many of the CTO’s and other chiefs admitted that the cloud offered more security and more accountability for their assets than hosting media locally. The control and scalability that cloud hosting offers makes sense from a business perspective, but a newly developing attitude of acceptance and preference of the cloud is a welcome change.

Live IP at BVE 2018

A distinction needs to be made between general use of packet switched networks, such as in head-end systems or post-production houses and live IP, which is a hot trend right now. Whenever IP is concerned, there tend to be two main camps. Those that see IP as a waste of investment, who do not trust it, and see no place for data centre hardware in broadcast networks. The other camp see an opportunity for growth and improvement, and see a future in which IP technology is a fundamental pillar of any operation. Some sit somewhere between the two. Five years ago, the first camp was dominant, but with every quarter that rolls past, the balance seems to tip towards the pro-IP camp. As live IP is adopted, and the benefits are demonstrated, more and more momentum will be gained by the second camp. SMPTE 2110 is by far the most serious standard in the space, and with it’s backing from AMWA/AIMS, it stands the best chance of gaining mass-market adoption.

The most serious use of live IP in a proven solution at BVE was the Timeline TV UHD-2, expanding outside broadcast truck. This multi-purpose OB facility has an IP core network, and is a shining example of SMPTE 2110 in action. It is capable of supporting HD, UHD, SDR, and HDR, making it extremely adaptable to customer needs. The most impressive gain that IP gives the truck is the density of its IP switch, and the associated savings that can be made in copper infrastructure. The modular Arista switch that sits at the core of the network utilises 100Gbps ports, making it roughly equivalent to a 4000×4000 HD-SDI router, but in 7 RU, and with space for expansion! The mind-blowing data throughput it supports allows the switch to support routing of the entire production. By using hardware end-points, narrow senders can be used, allowing for smooth traffic flow through the system. The user experience for vision mixers and engineers is the same as an SDI system, something that would likely have been deemed impossible 5 years ago.

IMF comes back into the spotlight

After it’s release some years ago, the Interoperable Mastering Format has seen a steady uptake from some fringe members of the industry, and notably – Netflix[2]. Bound by the constraints of legacy support for ingest of tape and files, broadcasters have had little motive to move to IMF, however, the case for doing so is now very compelling. When added to the decision made this week by the DPP, EBU, and NABA, which is underpinned by SMPTE (as author of the standard) to release a specification for IMF for delivery to broadcasters[3], now there is much more serious motive to support IMF. This is also very important because of the dominion of the organisations involved. It is a good and reassuring sign of international agreement and cooperation that the industry needs more of.

IMF was brought up in many of the sessions at BVE, and there is now enough awareness of the standard that end users are beginning to adopt it, with Endemol Shine group, Viacom, the BBC, Vice Media, and many more vying to introduce or continue introducing IMF to their work flow.

Quieter trends

VR & AR (or XR)

Although the creative stages had features on VR and AR, there is not enough movement on the technology to be able to offer these as services over-the-top of normal content. While XR is gaining traction in the gaming industry, it is failing to gain momentum in the broadcast and media space. Perhaps the two industries will better align in 2018, but to paraphrase one of the speakers, there are currently not enough standards to be able to act on the promises of the technology.

8K & “Super UHD”

Two years ago, the trend of “better pixels” was the hot topic in panel discussions. However, slow deployment and lack of consumer demand has kept resolutions beyond 4K on the back burner, at least in Europe. Until consumer demand catches up with the technology, SUHD may not gain much further traction.

Quantum Computing

Quantum computing is an exciting technology has been in the mainstream sphere for some time, but as far as I am aware, made no features at BVE. This could feature at NAB or IBC 2018, as big developments are being made, and quantum computing could become commercially available in the next 2 years. As the broadcast and IT industries continue to converge, such a large evolution of computing will surely have a large impact. The speeds achieved with quantum computing (using qubits), could see enhancements to AI that are unimaginable. Training of AI models could be done as fast as on the fly, to deduct patterns from data by analysing swathes of data in parallel. Google, an ever-growing giant in just about every field, is expected to have mastered quantum computing in 18-24 months[4].  As a cloud provider, and purveyor of AI technology, Google could come to dominate the AI offerings for broadcasters in the same way AWS does for cloud offerings.

What’s next?

If there was anything to learn about “the bigger picture” from BVE, it was that by taking a retrospective look at the industry, the 2016-2018 period is truly proving to being an extremely transitional period for the industry, perhaps more than originally thought. 5 years ago, the changes to the broadcast industry were largely internal, through development of new technology. Now they are external, and larger than ever. Mergers and acquisitions are likely to continue, and players that cannot adapt fast enough will be left behind.

How can BCi help you to leverage these new trends?

BCi is an established systems integrator with solid cloud experience, and future thinking experts that span the entire end-to-end workflow. BCi is actively challenging the distribution model for DCP packages in conjunction with Canadian cloud giant Cocego Peer 1, and utilising cloud services for transcoding with it’s Chameleon software transcoder. In addition, BCi is involved in leading research into the applications of AI within the broadcast and media space, putting it a step ahead of other players in the space.

Want to know more?