The New Siri? The Next Generation of Home Assistant Robots - LEKULE

Breaking

2 Jul 2016

The New Siri? The Next Generation of Home Assistant Robots

Home robots are becoming a popular product for businesses and independent developers to offer. Get to know some of the newest home robots.

Home Robots vs. Voice Assistant Devices

Voice controlled assistants started with the likes of Apple's Siri. Since then, Microsoft's Cortana, IBM's Watson, and Google's Okay Google have also gained popularity.
The first game changer was Amazon Echo, which is useful for both home automation as well as voice controlled computing.
But home robots have begun appearing on the market, introducing what could be the next step in smart home technology evolution.


Voice assistant logos, clockwise from the top left: Siri, Alexa, Cortana, and OK Google.

I know many would say “yeah, but I already carry my smartphone everywhere with me. Why would I get a home robot when I can use ‘Hey Siri’ or “OK Google’ for voice commands?”
Well, the idea originally behind Echo was that having to grab your phone for everything distracts you from everything around you, and using solely voice commands helps to alleviate this device obsession.
Personal robotics is an upgrade to the Echo concept. In the case of home robots, your device can follow you around, which means it is even more effortless and natural to interact with than Echo. Plus, the simplicity of the platform makes it perfect for kids or seniors who might have trouble giving complex commands like those often required to get exactly what you want from Google, let alone Siri.
Personal robots also assist with taking photos, videos, and video calls by tracking faces and always pointing the camera right at you.

Meet Zenbo

The most recent in a growing line of home robot releases, ASUS has taken home assistance several steps further with Zenbo.
From the Zenbo webpage:

"Zenbo is a friendly and capable home robot designed to provide assistance, entertainment, and companionship to families and meant to address the needs of each family member."
 

Image courtesy of ASUS.

Zenbo is clearly designed especially for the elderly and children. He has effectively all of the same features as an Android tablet, plus the ability to control IoT devices like smart TVs, smart lights like Philips Hue, and much more.

Like all other voice assistant devices to date, Zenbo can hear and respond to a large variety of commands, including ‘follow me’. He can play music, and dances along to it while doing so. Since he has a touchscreen, he also implements intuitive tablet-like controls, again like any other Android tablet. Like a voice-activated assistant, Zenbo can read messages and recipes aloud, but with the addition of the display.
But he has some important extras built in.


One of Zenbo's emotive responses. Image courtesy of BGR.

One of Zenbo’s best features is fall detection, which can automatically alert family members if someone falls and may need help, something that Amazon Echo certainly cannot do. He does this by detecting his surroundings, without the need for a vocal call for help. Plus, his ability to follow you around the house makes him much more accessible than Echo, which you would need to carry with you (or buy a bunch of Echo Dots for every room).
Zenbo also can act as a playmate for children, reading and interacting with them during stories (though it appears there's only one story available for now). However, that’s sure to change, as ASUS is opening Zenbo freely to developers! In my opinion, this is an awesome approach, as much of the interactivity that evolved/is evolving from VR is largely in the hands of dev teams, and giving this kind of physical interactivity in the real world to developers will open up yet more possibilities.


Image courtesy of ASUS.

For more information, here's an interview with ASUS chairman Jonny Shih about Zenbo.
Below is a demo video by ASUS to show everything Zenbo can do. (Be warned: YouTube commenters agree Zenbo may be the best actor in the video.)


Meet Jibo

Meet Jibo. Like Zenbo, he has a display as his ‘face’ (though he isn't capable of emotive expressions like Zenbo), can track faces, takes pictures, videos, and video calls while tracking faces, and so on, all just like Zenbo. However, Jibo is stationary. Also, while Jibo can interactively read stories to kids and use IoT controls through voice commands like Zenbo, he has no fall detection.



However, Jibo did get his debut before Zenbo on Indiegogo and with $3.7 million in crowdfunding, he certainly shows promise. His voice sounds considerably more natural, as Zenbo has been criticized for lack of natural vocals already.
However, on the feature side, both home robots seem extremely similar save for Zenbo's mobility and fall detection and Jibo's advantage of natural speaking. Jibo was funded in September of 2014, but like so many crowdfunded projects before him, has been taking forever to see a release and has only recently opened up for preorders. One great thing they do share, however, is an open SDK for developers.


Meet... Autonomous?

Meet.... actually, this one doesn't seem to have any sort of cute name. I find this odd, since it's referred to constantly as a 'she', much the same way Zenbo and Jibo are referred to as 'he' in their respective discussions and videos.
However, it seems you can choose whichever of several animated characters appears on screen, presumably with vocals to match.


Some of the faces you can select with Autonomous' home robot. Image courtesy of Autonomous.

She is marketed as the first Personal Robot for the Home by an already established company known as Autonomous. This might seem familiar to some; that's because this robot hit the news in early 2015 under the name Maya by a company known as Robotbase. Since then, Robotbase changed their name to Autonomous and seem to have ditched the Maya product name.
Autonomous is unfortunately not being as open as ASUS or Jibo and isn't offering developer tools to just anybody. Jibo is the company that came up with Jibo, of course, as they have emerged solely as the result of their huge crowdfunded start, whereas ASUS and Autonomous had their own funds to use for development. Strangely, Autonomous created a Kickstarter that raised over $160 thousand from what seems were essentially pre-orders (there were no backing levels for under $995 available and each gave the backer a personal robot).
That said, despite Autonomous getting this extra funding a few months after Jibo got theirs, they are actually accepting orders right now for their robot, and it has a set date (August 31st) when it should arrive too! But, since the Kickstarter, the price point flew all the way up to about double that of Jibo and a heck of a lot higher than Zenbo. So, what do you get for all that extra cash?
Autonomous has actually been way more open about their design than either competitor, going so far as to give out a facts sheet that details the hardware used in their robot.


The Autonomous team at work. Image courtesy of Autonomous.

Her design appears to be a mix between a Roomba and Double telepresence robot. Like Zenbo, she is mobile. Her awkward height means she can't move quickly without toppling over, so her top speed is a blistering 1.6 miles per hour.
However, she packs a bevy of sensors and features that either competitor lacks, like room temperature, humidity, and oddly even air quality (supposedly determined through a CO2 sensor) to assist with integration with a Nest or other smart thermostat. In the video, she demonstrates superb natural language communication, worlds away from Zenbo, and seems to be designed for far more than just home use, being an office assistant and telepresence robot as well as a home companion like Zenbo and Jibo.
However, all three robots are Android-based, function as video communication and camera devices autonomously, can help you keep track of your schedule, order food, talk through recipes... pretty much everything that Jibo can do, Autonomous' robot can too, and it seems everything except fall detection from Zenbo was added and then some.


"Adding Z-Wave and Zigbee." Image courtesy of Autonomous.

One thing that sets Autonomous' robot apart is the higher level AI of their machine. Their robot isn't simply programmed to adapt to its users in any simple manner. It actually utilizes deep learning powered by NVIDIA Tegra. It doesn't simply respond to orders and commands like the other two— it seemed in the video to make independent decisions without orders.
The video showed a number of such behaviors. The robot unobtrusively enters conversations in a natural manner to give reminders, and in the video seems to choose on its own to buy lunch at one point. Plus, when reading stories to kids, it can apparently manipulate the lighting of the room along with the story. If this product is everything the video makes it out to be, it could certainly be an advancement far enough ahead of Zenbo to justify the extra cost.



Everything said and done, all three of these robots are yet to be in use. I'm sure we'll find out for real what these guys bring to the table once they all arrive in people's homes! Who knows, personal robots could become as diverse and ubiquitous as their Android-powered cousin—the smartphone.

No comments: