Project Starline, a "holographic" video chat apparatus is back in the news recently. Is this a new version of telepresence? In this blog Bryan, a former developer of telepresence systems, shares his take on Project Starline and telepresence in general.
Commercially available telepresence dates back to the late 1990's. This was when bandwidth was expensive or simply unavailable to remote workers. Price tags on systems approached a million dollars, needed remotely managed and many rooms required extensive remediation. Big costs that only the Fortune 500 could support. Systems were designed to highlight "point to point" use and conveniently ignored remote workers (and multi-point calling for that matter).
Then bandwidth became more available, prices dropped and video conferencing could be used in typical meeting rooms. At that point, remote workers on a video chat when meeting rooms were present became the "odd person out", relegated to a tiny window and overall had a poor audio and video experience. This is due in large part to vendors putting their effort into making the meeting room experience great at the detriment of remote user - they wanted to simulate telepresence - only cheaper and more available. Well, we know what happened in the spring of 2020.
People working from home since 2020 are now in the position where they are having a better experience in video meetings than those in conference rooms for a variety of reasons. The display is closer, each person has their own frame, eye contact is better (for the most part), and we don't have to go to the office. In addition, there are excellent audio and video products available from vendors like Poly and Logitech to further enhance the experience. Understanding that workers aren't going to be flocking black to the office any time soon has forced meeting service vendors to reverse course and make the conference room experience as good as the remote worker has it...
That was a long preamble.
Google's Project Starline is a curiosity to me. It's a quasi-combination of three seemingly obsolete technologies - telepresence, 3D capture/display and a large apparatus for a video call. I reviewed this paper to get a better understanding of the technology behind the system. The system contains three sets of stereo cameras, one above left, one above right and one in the middle below, encoded and sent to the remote location. The remote sides splits and displays from the viewer's left and right eye locations (this means that two people cannot be in the same system using it). A significant note states "To account for this large parallax [location of the cameras to the user's eye line], we reconstruct geometric approximations of the user..". Therefore, this a recreation of the person.
The question is not whether we need/want high quality video conferencing. We do. The question therefore becomes: Do we need a multi-camera, 3D system usable for point to point only meetings? While the former telepresence designer in me loves that Google is working on this, the realist in me says no. Project Starline may be the next telepresence, but we don't need it to effectively meet.
Telepresence excelled for a short period in the days before we could have a video chat from just about any commercially available smartphone or laptop for free. I don't think it's due for a comeback. We used to talk about the benefit of telepresence being that a person didn't have to travel to be in a meeting. We have that now. We can meet with anyone from the comfort of our home office, coffee shop or back porch. Any telepresence system will require travel to an office as they won't be installed in anyone's home. I'm all for a high quality, face to face meeting. But we have that. From home.
This blog was based on information attained via publicly available material. Neither Bryan, or the UC Test Lab has received a demo or any additional private information about Project Starline.