Cisco Gives Glimpse of Collaboration FutureCisco Gives Glimpse of Collaboration Future

Siri, what's the weather today? Alexa, is it cold outside? OK Google, do I need a jacket this morning? Weather status is the most common voice inquiry asked of consumer-oriented virtual assistants, as Cisco Collaboration chief Jonathan Rosenberg remarked to press and analysts gathered at Cisco's San Francisco Innovation Center this week. But when it comes to virtual assistants in the office environment, things aren't quite as simple.

In fact, designing AI systems for enterprise and business user requirements is quite complex, as I learned at the half-day event showcasing how Cisco is bringing artificial intelligence (AI) into its collaboration portfolio and beyond. As such, Cisco intends to be "pioneering [in] this field for many years to come," Rosenberg said.

To help guide its AI efforts, Cisco surveyed more than 2,000 workers on attitudes toward AI (see "Meetings, Meet AI"). "We wondered, if we build it, will they come? And thankfully, the answer is yes," he added, noting that 95% of respondents indicated they would use AI for work tasks.

The Rise of Ambient Computing
Computing is becoming ambient, "embedded into everything around us" and opened for access in shard spaces like our living rooms at home and conference rooms in the office, Rowan Trollope, SVP and GM of IoT and Applications at Cisco, remarked in opening the event.

But four key drivers make ambient, shared computing in business dramatically different than what's at home. These are identity, context, environment, and security, Trollope said. Take identity as an example; a virtual assistant in your living room only has to worry about identifying people who live in the house, but a virtual assistant for your office has to deal with identifying hundreds of employees that share workspaces.

To solve the identity problem, Cisco invented an ultra-high frequency pairing and identity technology for the virtual Spark Assistant it introduced last week. The technology incorporates mobile devices, facial recognition done through vector analysis, and speaker tracking performed through its video and audio equipment. Once Spark Assistant has confirmed a user's identity, it will be able to provide meeting details and other contextual information as appropriate. Needing to be able to deal with the context of teams is one of the reasons Cisco acquired AI platform provider MindMeld earlier this year, Trollope noted.

In a modern meeting space, what's on an individual's smartphone "moves from the hand to the wall," requiring security and environmental considerations that don't come into play for home-use virtual assistants, Trollope added. For example, you wouldn't want the Spark Assistant to expose your personal or confidential information to all meeting participants. But if you're alone in a conference room with a Cisco room system, Spark Assistant will recognize that and display content that's appropriate for your eyes only.

But when it comes down to it, Cisco's efforts are "all about how we help teams get work done together," Trollope said.

Click to page 2 for more on what Cisco is doing with AI today and in the future.Continued from Page 1

Cisco AI: What's Being Done Today
My fellow attendee, UC analyst and Cisco watcher Zeus Kerravala, told me this one-day Cisco AI gathering was one of the most visionary UC events he has attended. "What Cisco showed is how machine learning can be used to change the dynamics of a meeting to make everyone significantly more productive."

So how will AI changing the dynamics of meetings? This first version is a voice assistant that automates call initiation for ad hoc or scheduled meetings and call control tasks, and identifies people in a collaborating session. A lot goes on behind the scenes to accomplish these items.

For example, Cisco has enhanced its video codec with AI to improve speaker framing in video conferences. The system will recognize that only two people are in a room, and frame the view on them, eliminating empty space. Should a third person enter the room, the system would widen the view to add in that participant. The codec learns framing requirements throughout the meeting, making adjustments more quickly and accurately as the meeting progresses.

In addition, in-room Spark video systems can detect and suppress background noise, so the days of having meetings interrupted by a colleague's barking dog are long gone. In fact, Rosenberg quipped, Cisco even has data and sound heat maps for 100 different breeds of dogs -- so you can rest assured that whether you have a Pomeranian or a Labrador, Spark will suppress your dog's barking.

To identify meeting participants, Cisco uses facial vectors that map the distances between nose, eyes, and chin to create an array unique to each individual. So as you can see in the picture below, when attendees lined up to see the demo, the system recognized who we were and added name badges below our images. (That's me taking the picture on the left.)

"Given how long we have been conducting meetings, it's fair to say that we have yet to perfect the art of running them efficiently because there are too many tasks filled with human latency," Kerravala said. "By embedding machine learning into the meetings, Cisco is having machines take care of the onerous tasks we aren't good at and also improve the quality of meetings."

Cisco AI: What's Coming
Beyond what AI can do for meetings for today, Cisco showcased some of the innovation coming out of the Emerge Lab, which works on figuring out the future of AI in teamwork and collaboration. One of the demos the Emerge team showed was a beta product, called TeamTV, that essentially turns Spark team spaces into live, always-on TV channels.

TeamTV is aimed at capitalizing on the informal conversations that help build trust and foster creativity -- the "unmeeting," in Emerge Lab parlance. As you might guess, AI is woven into this product in a number of ways. The Spark room system recognizes gestures to bring remote team members into the space. For example, a user can wave at the system to get its attention and invoke an action, such as muting or blurring a team member's video image. And it uses facial recognition to introduce team members to one another. But one of the cooler experimental features is an always-listening AI-enabled virtual assistant that can proactively contribute to conversations. For example, if a team is talking about a particular company, the system will pull up that company's profile and stock price on the screen. This is just another way that Cisco is turning AI into an active meeting participant.

The Emerge team also showcased Cisco Spark in VR, which is now available in the Oculus store. The AI assistant within Cisco Spark in VR can do things like build a customized virtual meeting space and populate the room with 3D images. But it can also learn and bring in knowledge and resources. In a demo, the team prompted the Spark Assistant to learn about jet engines and after a quick scan of Wikipedia, it was able to answer questions like how much a jet engine weighs or what causes jet engine noise.

Of course, big companies dabbling in advanced technologies like virtual assistants, AI, and virtual reality have a responsibility to figure out the difference between what they can do and what they should do -- and Cisco takes that to heart, Trollope noted. "We have a very big group of people that dedicate themselves to legislative and privacy concerns so that the 'big brother problem' doesn't become a reality. If a virtual assistant is spying on you, no one is going to use it. The symbiosis of man and machine is where we are going."

Hear directly from Rosenberg on Cisco's vision and product direction in a keynote address at Enterprise Connect Orlando 2018, coming March 12 to 15. He'll take the stage on Tuesday, March 13, at 10 a.m. Register now using the code NOJITTER to save an additional $200 off the Advance Rate or get a free Expo Plus pass.

Related content:

Follow Michelle Burbick and No Jitter on Twitter!
@nojitter
@MBurbick