The Telepresences Long View

As corporate America becomes more acquainted with telepresence, innovators already are looking at ways to evolve the experience, conducting pure research with an eye toward the sector years from now.

One vendor, Teliris, recently announced it was untethering its research lab, created more than a year ago, as a discrete unit within the company, decoupled from profit and time-to-market imperatives. The idea is to free up engineers to dream up applications and develop technologies that won't necessarily bear fruit for years but, when they do, could spur shifts in user behavior. Teliris is adding staff and funding to the lab, which has almost doubled in the last year to more than 20 people.

“We have our own cost center now, and we have our own physical space here in the office, which we don't let the CEO into that often,” said Steve Gage, executive vice president and co-founder of Teliris and head of the labs.

Among the areas Teliris is looking at for long-term potential are linear photography as a way to divide a telepresence room into more segments (ultimately perhaps infinite segments) for transmission purposes. Teliris' system today has four cameras, each with its own codec, which imposes some limitations. A person sitting at one end of the table might have good eye contact with his counterpart on the screen directly across from him but poor eye contact with the person onscreen at the table's opposite end. Having more codecs won't help because transmitting multiple codecs simultaneously takes a lot of bandwidth. Linear photography could break the room into more segments practically in a multipoint session, but Teliris says it could take two to three years to figure out how.

Another area is higher-density sensors for multipoint calling. Today in a three-way call, cameras zoom out to include both of the other parties. With a larger digital sensor, however — one larger than 16 megapixels at 60 frames per second — Teliris imagines being able to do that zooming in the digital domain rather than using optics. That will probably take more than two years to achieve, Gage said.

In the shorter term, Teliris will explore the potential for biometrics and pattern recognition (identifying users' fingers or faces) to replace the RFID technology of today. Pattern recognition could help automate the reconfiguration — made manually today — needed when new users enter the room. And it could treat users accordingly in terms of preset policies related to business logic, security and functionality.

“If you have a document on the table, that may not be one you want the person in London to see,” said Mack Treece, president of Teliris. “By having rules based on persona, you could not pass the media across the table to them in London. It will not physically go across the table.”