Telepresence on the horizon?

As Andrew Davis and Ira Weinstein, Wainhouse Research, report, the ripple effect is already being felt, but the long-term results are anything but clear.

Telepresence has been in the news lately. The resurgence of articles began in early 2006 with HP’s announcement of its Halo system – a telepresence solution designed in conjunction with DreamWorks. A few months later during Q2-06, Polycom joined the fray by announcing its RPX product line of telepresence suites with expected deliveries to begin in Q4-06 or Q1-07. Around the same time, Cisco’s CEO, John Chambers, began talking about a telepresence system that his company would introduce before year-end. While no details were forthcoming, the Cisco pre-announcement created fear, uncertainty, and doubt in the minds of videoconferencing suppliers and buyers alike, while also creating new interest and excitement. The uncertainty factor was eliminated on 23 October 2006 at the global launch of the Cisco TelePresence 1000 and 3000 systems. The combination of announcements and press coverage, punctuated by the Wall Street Journal story on 28 September 2006 has generated more customer inquiries at Wainhouse Research in a shorter period of time than any other development in the history of the videoconferencing industry. Never before have CEOs been as interested in videoconferencing as they appear to be today!

But despite the hype, the fact is that telepresence is not new. One of us saw our first videoconferencing telepresence demonstration in 1995. Multiple vendors have been in this market for many years, but none of them have achieved any scale. A fundamental business obstacle is that these suppliers are in the position of asking major corporations to buy very expensive and totally proprietary systems from tiny vendors in weak financial positions. It is no wonder that telepresence has achieved little market penetration.

In search of a definititon

It only adds to the confusion that there isn’t any accepted industry definition of telepresence. Many people ask Wainhouse Research what the difference is between videoconferencing and telepresence. We believe there is a fundamental difference in the environment and the sensation of immersion that distinguishes a telepresence solution.

Our definition: Telepresence = A videoconferencing experience in which the participants feel like the remote participants are in the same room.

The telepresence design envelope

In designing a telepresence solution, engineers have focused on three factors they believe are critical to achieving customer acceptance:

High quality audio and video

In order to provide the realism users expect in a telepresence meeting, the audio and video must be clear and with no noticeable delay. Realism is optimised when the images are lifesize, when the gaze angle between the camera and the remote video is minimised, and when the audio provides directional cues as to who is speaking. The face-to-face experience is also maximised when the technology is hidden; i.e. cameras, speakers, and microphones are nearly invisible.

Simplicity

The intended audience for telepresence solutions is CXOs and senior level managers with little interest in working complicated communication devices, and even less interest in looking clueless in front of their peers when trying to set-up video calls or fix problems. For this reason, telepresence solutions must be extremely easy to use (one button to push) or offered as a totally managed solution (theoretically with zero buttons to push).

Reliability

When groups of senior managers meet, even short delays to get the call running can be frustrating as well as expensive in time wasted.

Many telepresence systems go one step further with the vendors being heavily involved in room design. When the chairs, tables, background, and even the color of the walls in the remote room are the same as those in your room, it is easier to get the ‘telepresence feeling’ that the remote colleague is indeed in the same room with you.

As is the case with any engineering design, compromises are involved. Telepresence systems designers have had to yield on cost. Most use proprietary audio-video algorithms to maximize quality, resulting in high bandwidth requirements and little interoperability. Most of the systems in the market today, in fact, cannot communicate at all with systems from any other vendor.

Any enterprise considering a telepresence investment should consider carefully many factors surrounding the vendor (stability, customer references, support capabilities) as well as around the solution. Five of the top technical factors might be:

Video quality

Many but not all of the systems today are moving to HD, providing superior detail compared to standard definition television. Some compression algorithms have been optimized to reduce delay, a key parameter to enabling ‘natural’ conversations. Many people feel it is important to the experience that remote participants appear life size, but not all systems can do this.

Eye contact

If the distance between the camera lens capturing your face and the eyes of the remote participant on your local display is relatively large, then the remote participant will have the feeling that you are not looking directly at him during the meeting. This is known as the gaze angle problem (angle can be horizontal or vertical), and a significant gaze angle can significantly detract from the meeting experience. Eye contact instills trust and fosters an environment of cooperation and partnership, while the lack of eye contact can generate feelings of negativity, discomfort, or distrust. Providing natural feeling eye contact during a videoconference requires that the participants look almost directly into the camera. Unfortunately, videoconferencing often fails in this regard because participants tend to look at the video image of the remote person and not at the camera.

Audio quality

Wideband audio provides a significant improvement over narrowband audio, increasing clarity dramatically while decreasing ‘meeting fatigue’. Separate from audio bandwidth is the issue of multi-channel audio. Multi-channel audio combined with multiple speakers contributes to the realism by providing spatial hints about the speaker’s location, but is technically complicated because of the need for multi-channel echo cancellation. Different systems vary widely in their audio capabilities, and we recommend that users experience the sound as part of their decision making process.

Bandwidth requirements

Most traditional business videoconferencing today takes place at a connection rate of 384 kbps, a legacy from the days of ISDN communications. While many companies have migrated to IP video, few conduct videoconferences over IP at more than 768kbps. Telepresence systems today, in contrast, typically run over dedicated IP links and bandwidth requirements range from 1.5Mbps to over 24Mbps per session, depending on a variety of factors. The result is that few IT departments can (or want) to run telepresence over the corporate IP network; instead, an overlay IP network strategy is preferred. This can add considerably to the monthly operating costs.

Multipoint support

Not all meetings are point-to-point. Sometimes conferences involve 3 or 4 sites, or even higher. Not all telepresence systems can support multipoint operations.

Four other distinguishing factors between vendors’ solutions might be considered under the category of user preferences including:

Dedicated room designs – Several systems on the market are delivered in a ‘room within a room’ form factor, while others are designed or installed in standard conference rooms. Which design type is better suited for your organization depends upon your environment and budget constraints. Furthermore, some designs use special tables and room layouts that work well for telepresence sessions, but are far from ideal for in-person meetings.

Bundled services – Some telepresence solutions may bundle in a variety of services including equipment installation and maintenance, call launch and management, and network services. Customers need to consider how the telepresence suite may fit in with their existing network provider contract, conferencing managed services organization, and/or internal IT support team.

Interoperability – Maintaining all the elements needed to support telepresence generally means limiting communications to telepresence rooms from the same vendor. But some systems will support connections to the industry-standard (read H.323) systems that have been widely deployed for the past decade. True, connecting at low bandwidth to a standard system will sacrifice the telepresence experience, but this situation may be preferred to not being able to connect at all.

Price – Typical group videoconferencing systems today (codec, camera, microphone, speaker, display, cart) range in price from US$8000-18,000 (US MSRP), with some of the high end systems approaching US$28,000. Telepresence rooms, on the other hand, range from US$80,000-400,000 and up depending upon whether extensive room enhancements are needed prior to installation. Because of their high up-front costs and operating costs, telepresence systems are likely to remain an industry niche suitable only for large enterprises for quite some time.

Telepresence market impact

The entry of Cisco and HP into the high-end videoconferencing business has certainly had an impact on the market. With their access to C-level executives, HP and Cisco have already caused a shift in the ‘decision making unit’ for visual communications. Wainhouse Research has fielded numerous calls from videoconferencing managers telling us: “My CEO saw a telepresence sales pitch and is asking me why we don’t have one of these.” It appears that telepresence decisions are being made by the same team that selects the corporate Gulfstream. We suspect that 6-9 months from now many video resellers, typically selling to videoconferencing management, may visit with their key accounts and see find themselves confronted by fully installed telepresence systems that they had no idea were even being considered.

Telepresence solutions will match certain niche market opportunities well. For example, an enterprise with a small number of headquarter sites (NY, London, Tokyo) conducting frequent meetings between high level executives in those sites will find telepresence an attractive alternative to airline travel. In addition, those involved in telemedicine and education may find that the telepresence experience changes the learning environment. But any enterprise with dozens of remote offices, manufacturing plants, or partner operations will likely find the high initial cost and limited interoperability to be major stumbling blocks to serious consideration.

The truth is that telepresence shows just how good a videoconferencing experience can be. And the Cisco and HP marketing muscle is likely to raise the fortunes of many involved in telepresence and high end videoconferencing. We believe that many savvy customers will find that a knowledgeable systems integrator using a high-definition videoconferencing system from LifeSize, Polycom, Sony, or Tandberg may be able to deliver the vast majority of the telepresence benefits at a small fraction of the cost, while also providing interoperability with installed systems.

Despite the hype, telepresence does not change the four long-standing elements needed for videoconferencing adoption:

• Voice quality equal to or better than the telephone.
• Video quality that adds to the communications experience, not detracts.
• Strong ease-of-use and reliability, thereby eliminating the need for in-room support staff.
• Interoperability and ubiquitous access that allows a video site to call many other video sites.

In short, telepresence should drive interest in videoconferencing overall, improve the efficiency of a lucky few users, and drive market growth for many within the marketplace.

Andrew Davis is the Managing Partner with Wainhouse Research. He can be reached at andrewwd@wainhouse.com. Ira Weinstein is an Analyst and Partner at Wainhouse Research and can be reached a iweinstein@wainhouse.com.

Wainhouse Research (www.wainhouse.com) will be publishing a full report on telepresence systems during Q4-2006.