Next-Generation Telepresence Is Here And It Works

It's been a long time since any claim of a "next-generation" technology has held up against our twice-bitten mentality, hardened by years of IT hype. But telepresence appears to be succeeding.

These systems combine room-sized audio, video, and ergonomic design to simulate live interaction. They're the stuff of George Jetson and Buck Rogers, filling a gap in the communication options available today.

But the systems don't come cheap. They start at $30,000, with a fully functional room that seats eight to 16 people hitting $300,000--per site. The high-def cameras, video screens, and integrated sound all drive up the cost, as does the infrastructure and network upgrades required.

Major vendors include video pioneer Polycom, as well as Teliris and Vidyo. Cisco is a force, having fleshed out its video communications product set with its acquisition last year of longtime player Tandberg. Mouse and desktop camera maker Logitech raised eyebrows when it bought LifeSize last year, a videoconferencing and telepresence system vendor.

Telepresence systems combine updated compression and video protocols in high-def TV with good old-fashioned human factors engineering to add the subtleties of personal conversation--eye contact, voice tone, body language--to traditional video. These systems don't have a broad range of interoperability and have bandwidth requirements that make Web-based usage questionable. Nevertheless, folks who are using them swear by them.

What's Required?

A true telepresence system that provides a near real-life experience has five requirements: high-definition, life-size video; camera technology that simulates eye contact; high-quality sound that allows movement and multiple conversations; a room design that simulates an in-person meeting; and enough bandwidth to avoid jitters and screen jumps.

All systems that you connect in should meet these requirements. If they don't, users on the lower-quality system, at best, will end up only able to passively listen in. At worst, they could drag down the quality of the entire system. It's not unusual to have two sites collaborating fine, but add a third, below-par site and there's a pixilated jerk screaming "What'd he say?" every five minutes.

"A telepresence system is more akin to a new building project than a technology initiative," says Bill Galinsky, CA's VP of global infrastructure. CA is a longtime user of videoconferencing systems, with more than 2,800 units in place--everything from desktop video to telepresence.

To get a sense of the scope of these telepresence setups, think of building a data center that will be open to the public. You need to get your facilities group in lockstep with your infrastructure, server, and application teams, and make sure all the rules about usage, bookings, and meeting length are defined. Start early and identify any infrastructure or bandwidth gaps that need to be addressed. Focus on the entire video and voice path, everything from core switches and routers to your MPLS network.

It's A Private Cloud

Time to dispel one myth: You're not going to use a true telepresence system over the Web. To their credit, most vendors have done a good job outlining their bandwidth requirements and recommend private clouds for communication. However, they all tend to dance around the Internet issue.

Unless you've planned for future video bandwidth during a previous upgrade, it's doubtful you have a large enough private pipe. How much do you need? Start at 5 Mbps for a small room and jump up by increments of five for every screen in the room. Too conservative? See what happens when you drop half a million dollars on a project only to have jittery connections.

Plan to use telepresence systems internally or with a small, well-defined group of customers. Don't count on being able to easily connect with anyone else in the future. Integrating different videoconferencing systems has always been difficult. Hitting the performance and user-experience benchmarks expected with telepresence is years away.

It's not that vendors aren't willing. Polycom has long based its systems on some core standards. Cisco has submitted its telepresence protocol to the Internet Engineering Task Force and announced Media Experience Engine 560, which will support translations from Tandberg, Polycom, and LifeSize. But interoperability continues to be out of reach.

Being able to tie in lower-end systems is also important. Most telepresence systems already have some basic connectivity, and Cisco has even announced basic WebEx connectivity. But not everyone is buying into the usefulness of doing this. "You need high def to work with high def. Tying in lower-end systems ends up being annoying for everyone," says one Cisco telepresence customer who wasn't happy with how his systems worked after his boss insisted on linking in lower-end systems.

If You Build It, Will They Use It?

Traditional videoconferencing systems have a checkered history of sporadic use. One global telecom firm that asked not to be identified recently discovered average videoconferencing utilization rates for its systems to be around 16%. Ouch.

A key factor is making telepresence available to everyone, not just senior managers. "We opted to give all users access, plus the ability to book meetings themselves. The only restriction is availability," says CA's Galinsky. People simply book a system and show up. IT does the rest.

CA started its telepresence rollout in the fourth quarter of 2008 with nine sites and saw usage jump to 40% in the first few months. It attributes the high adoption rate in part to the fact that its business units weren't happy with the performance and complexity of existing videoconferencing options.

The ROI Conundrum

Telepresence saves a company time and money when it's used to cut the number of staff having to travel for audits, sales meetings, engineering reviews, and similar gatherings.

We looked at a simple scenario where two sites are connected via telepresence systems that have an estimated $115-per-hour cost of operation. We applied this scenario to a situation where telepresence didn't replace on-site work, but instead gave the organization the ability to send fewer staff members (one instead of three) to accomplish the same amount of work in the same amount of time (four days). In this case, the savings in transportation and hotel costs more than offset the cost of using the telepresence systems, providing an estimated $1,345 in savings.

This scenario underscores the importance of accurately predicting your organization's telepresence utilization rate. It assumed a utilization rate of 70%; if you have a lower utilization rate, the telepresence systems' cost per hour increases, cutting potential savings.

That's what a New England prep school's IT director found out too late after installing a Polycom system five years go. "It kills me to see that room sit," he says. "We could have done so much more with the money."

The school's system was part of a larger capital project. When the overall project faced budget cuts, decisions on what to keep were made based on the capital costs of each piece of the project. Had the school considered the operational expenditure side of the equation--particularly whether its utilization rate would yield sufficient operational savings--the decision to cut other projects might have been different.

The school's experience underscores the importance of considering both capital and operational costs when assessing a telepresence project. For some organizations, high utilization rates and greater operational savings could be key to getting the green light.

In addition, creating a point ROI model just for telepresence may work for some organizations, but a broader model that includes multiple communications options is probably better. Bring in all the options--telepresence, traditional videoconferencing, desktop Web conferencing--and you get a framework for a total cost analysis that will help you calculate the big-picture savings for your CEO and CFO. (Our full InformationWeek Analytics Report includes an ROI model that lets you test which options are best for your company. Get it at informationweek.com/analytics/telepresence.)

Bottom Line

If you implement a telepresence system as a siloed project, its high overall cost and relatively low level of interconnectivity may set you up for failure. Instead, think of telepresence as the top of the continuum for your communications vision. When done correctly, a true telepresence system is a viable alternative to some face-to-face communication. And its "wow" factor could help jump-start your broader unified communications effort.

Michael Healey is president of the Yeoman Technology Group, an engineering and research firm that maximizes technology investments for organizations. He can be reached at [email protected]