"The way that the personal robots revolution is going to really happen is by making it a platform," she says. "Because once you do that, suddenly you can have a robot that can do many things for you, for many different people, versus these niche robots that only vacuum or only clean your gutter."
Jibo is powered by an ARM processor and runs Linux. It has two cameras, which allow it to detect and track people, a microphone array for sound localization, and touch sensors on its body. Three motors and a belt system allow Jibo to move fast and smoothly. It uses WiFi for connectivity and will be typically plugged in for power, though with a battery (not included) it will be able to operate for about 30 minutes.
Here's Breazeal describing the robot and her vision during our meeting:
I mentioned to Breazeal that I felt some people would not call Jibo a robot. I mean, it's a robotic device, but is it really a robot? I'm sure some people will ask, "Where are the arms?" and "Why can't it roll around?" I asked Breazeal if she explored prototypes with arms or a mobile base, to give it more capabilities around the house.
Breazeal says there's no question that Jibo is a robot. "It's a social robot," she says. And no, she insisted that they did not build robots with arms or able to drive around. Because of their cost and complexity, manipulation and locomotion are "way, way down the line."
Breazeal is famous for the highly expressive robots she and her students built at the Media Lab, where she directs the Personal Robots Group. So prior to meeting Breazeal and seeing Jibo, I was trying to guess what the robot would look like. Would it have big eyes like Kismet? Fur like Leonardo? A tail like DragonBot?
After seeing Jibo's sleek, clean, Apple-esque industrial design shineness, I asked Breazeal if she considers Jibo a big departure from her previous robots.
She explained that the design choices were deliberate, and part of how she and her team envision bringing personal robots to people's homes. "By keeping it more abstract you’re also keeping it more sophisticated. And we want people to think about this not as a toy -- because it’s certainly not a toy -- but as this high-end consumer electronics device."
"The core competency of a social robot is its ability to engage people," she adds. "Not only just around information but around social, emotive, interpersonal dimensions. And even through its physical, social embodiment. That’s the core competency. is designed to hit that as powerfully as we can with the most elegant, sleek, minimal, sweet spot of that. There’s a kind of Occam’s razor thing going on here."
So, will people want it? I’d say yes, if Breazeal and her team can deliver a virtually flawless user experience. If Jibo works most of the time but not allthe time -- if you have to keep repeating yourself to be understood, if the robot fails to recognize you, if applications are slow or stall -- this will likely make it a big failure.
In fact, Breazeal says that the biggest technical challenge in developing Jibo is integration. "The sensing, the vision, how it all comes together to create this magical experience when it’s just you and your Jibo, that’s what the biggest challenge is."
To be sure, if anyone can deliver on a robot like this, it’s Breazeal. And she’s not alone, having assembled an impressive team with executives and engineers with backgrounds in speech recognition, natural language, user interaction, gaming, and animation.
Will Jibo be a big hit? Will it become "part of the family"? Leave your opinion in the comments below.
Get more from IEEE Spectrum
This article originally appeared on IEEE Spectrum; all rights reserved.