Case study: How CERN scientists collaborate with Vidyo

The European Organisation for Nuclear Research – more commonly known as CERN – is located in Geneva in Switzerland but is globally renowned for its groundbreaking scientific work, be it creating the World Wide Web or seeking out the "God particle" with its Large Hadron Collider.

But, despite having such a famous central base, CERN is actually made up by over 20,000 scientists from 113 different countries based at 600 universities and institutions across the globe, all collaborating with each other on numerous projects.

As a result, communication is key to the organisation and CERN chose very early on to use video conferencing , ensuring its members could speak face to face whenever they needed.

CERN took its first steps into video conferencing in 1996 when as a scientific organisation, it decided to develop its own system. 

“CERN often encounters technical challenges as we get there first, be it due to our bigger scale or more complex set-up, and we often need to develop our own solutions for such problems,” said Tim Smith, group leader of collaboration and information services within the IT department at CERN.

“In the early 1990s, with the collaborative nature of how we work, it was only natural that we looked to video conferencing and realised we needed face-to-face interaction with our physicists based around the world, so we became the first adopters and, indeed, developers of some of the technology.”

The collaboration team developed what Smith referred to as a “rudimentary” solution but he said it rapidly took off as the home-grown technology was tailored to their needs. However, even the best scientists in the world knew the solution would not last forever.

“As is often the case with early adopters, the world outside catches up and often overtakes us, thanks to both open source developments and commercial offerings,” added Smith.

“We were happily going along throughout the 1990s and much of the 2000s but in this time the world was racing along and new technology was being created with new standards and other new solutions.”

So, in 2009, the IT team at CERN began to examine other options and Vidyo came top with its use of SVC – Scalable Video Coding – technology, which enables high-definition video to be streamed using subset video bitstreams to reuse the bandwidth requirements of large video transfers.

In 2010, Smith and his team decided to implement a pilot of Vidyo, won over not just by its performance but by its interoperability.  

“One of the key things for us is it had to work on every device,” he explained. 

“We do not want to control nor do we have any desire to control the devices that our collaborators use, be it Windows, Mac, any flavour of Linux, mobile devices etc. However, we do want to be sure that quality is kept up across all those devices.”

The IT department took its time with the pilot though, taking almost two years before a full implementation.

“We gradually expanded the pilot and worked with the company to ensure it met our needs but we didn’t jump over straight away by any means,” said Smith. “For example, Vidyo was not mature enough when it came to its Linux integration initially but, by the end of 2012, we felt all our needs had been met and we considered a full move.”

We were happily going along throughout the 1990s and much of the 2000s but in this time the world was racing along and new technology was being created

Tim Smith, group leader of collaboration and information services, CERN

The old internally developed system continued to run in parallel with Vidyo until the end of 2012 and when New Year came round, CERN made the switch full-time to the new system.

“It was relatively smooth, with a few operational issues such as interconnecting all the gateways, SIP, and traditional phones, but when we switched over we were precisely ready,” said Smith.  

“Since then there haven’t been any major hiccups and the deployment and upgrade of clients was quick and smooth. One of the benefits of working with a commercial provider is when they put things together in that world, they think more of deployment and upgrades than we do when we are developing solutions in-house.”

The reaction of staff at CERN and the physicists using the system around the globe has been a positive one, and they are a tough group to please.

“We are no different than any other community in the fact our users want a smooth transition with no downtime of any new service. Unless the benefits are going to be huge, they won’t accept anything else,” said Smith.

“They are different to your average business in that they are very hands on, experimental and knowledgeable so if there are any issues to overcome they will probably try and fix it themselves. They are very nice in that sense and are always trying new things. But, as they know technology and need it for their everyday life, they are very demanding and have no tolerance for outages.”

The deployment of Vidyo and video conferencing at CERN is vast and has truly helped the community of scientists collaborate to address some of the world’s most challenging scientific conundrums. Yet video conferencing still has a relatively low adoption in other enterprises and the much touted "year of video conferencing" has yet to come to fruition.

So how does Smith envisage the technology’s future outside of CERN?

“In the early days of video conferencing it was all proprietary systems that were often very expensive and often didn’t interoperate with one another, putting up barriers to adoption,” he said. “With the advent of more and more web based technologies, albeit with the applications local, the technology became more pervasive and easier to use.

“There will always be different local requirements and quirks but as we push towards web-based solutions more and more with the same level of codec, I imagine adoption will rocket. We have gone through many waves and paradigms of computing, with more of a focus on HTTP. When we reach that stage with video conferencing, everyone will be familiar with it.”

The issue of quality and reliability has also been a factor in recent years, putting people off from trusting the technology, but again Smith sees this changing.

“You could always, no matter how bad the line was, make out what someone was saying over a phone call,” he said. 

“In recent times that is where video conferencing has fallen apart. But the latest codes adapt to hard conditions and you can rely on them now almost as much as you would a phone line.”

Smith concluded: “We are almost at that year of video conferencing that has often been sought, but it will be the web and the right codecs that will make it happen.”