Ubi – The Talking Computer Project

Ubi is a voice-activated tiny computer that plugs into a wall outlet and is programmed to answer questions and perform tasks based on simple voice commands. The Ubi connects to the Internet through WIFI and includes all sorts of sensors which can help it make sense of the environment and make many tasks currently impossible (or very costly and expensive), simple and straight forward.
–

Ubi is a Kickstarter project (crowed funded) and in the few weeks it’s been online it was able to rise more than $165,000 dollars from Kickstarter users who found the concept appealing and were willing to spend 150-190 dollars on the Ubi (surpassing it 36,000 dollar goal).
–

TFOT recently had a talk with Ubi Team Members: Amin Abdossalami, Mahyar Fotoohi, and Leor Grebler to learn more about Ubi and where it is heading.

The Ubi Kickstarter project from Kickstarter

–

A: The Ubi is a very simple computer that plugs into a standard power outlet and allows you to access the Internet through voice. It’s also a lot more than that – it comes with sensors that can analyze its surroundings and can send this information to a web portal. It also has lights than can be used as indicators for different events online.

We came up with Ubi as a shortened version of "ubiquitous computer". In essence, the Ubi is this type of computer. You plug it in and it just activates when you need it, blending into the background.
–

A: At the simple end of tasks, Ubi will allow you to search the Internet through voice and have answers read back to you. You can also make calls, access data about the room in which the Ubi is placed, and have the indicator lights turn on and off based on different events.
–

A: On delivery, we plan on allowing the Ubi to have: Internet search through voice, speakerphone, indicator lights based on triggers, music piping (music throughout your home), virtual assistant (calendar, news feed reader, email, voice memos, alarm clock), intercom, baby monitor, sensor system, and home automation system controller.
–

A: Yes! Absolutely, this is one of the most useful features of the Ubi. The answers will come from a Google search and will be read back.
–

A: A project like the Ubi wouldn’t have been possible even less than a year ago. We definitely aren’t competing with these giants who have armies of developers and engineers working on voice recognition. They do fantastic work. Instead, we’re extending the technology that’s already being used by creating a new form factor for accessing this ability to do voice processing.
–

A: Here are the specs of the Ubi:

A: We use a lot of the technology that comes with Android 4.1 Jelly Bean for doing voice search, such as the way Google Now reads back single line results, sends memos, or sets alarms through voice. However, there’s a lot more that needs to be done in order to activate these features without being able to see a display or touch a screen. Also, we need to determine what someone is intending to do based on their voice commands, like knowing which apps to launch and then being able to close these apps without requiring intervention from the person using Ubi.
–

A: We’ve always been really fascinated with the idea of the Singularity – the time when the processing power of computers will surpass all the minds of mankind. At this point, hopefully, there will be a kind of utopian reality when most of the ailments and problems of humanity will be solved. One of the technologies that we need to get us there is a computer that’s always on and ready to assist us.

The three of us are engineers (two of us have Master’s in engineering) and we were always throwing around ideas on what the future will look like in the next five or ten years. One of the ideas that kept on coming up was that computers will be less like the computers we use today – they’ll be more intuitive. Also, they’ll be less intrusive: available when we need them but then blending into the background when we don’t. The other big idea was being able to actually have something like the Star Trek computer in our homes. It was just under a year ago that Siri was released and everyone was talking about the Star Trek computer.
–

So with those items to motivate us, we started to work on the project and went through many iterations of design and concepts until we came to the form factor now being presented on Kickstarter.
–

A: We hope to ship the first beta versions of the Ubi in January 2013 as well as release the SDK around that time. Kickstarter backers should get their Ubis in February with the full release of the Ubi happening around March 2013.
–

A: The Ubi will have a trigger word for getting it to do what you want. Initially, that will be "Ubi" but we hope to make this changeable to other words.
–

A: We plan on shipping the Ubi with temperature, humidity, air pressure, and ambient light sensors. We can also feed back microphone volume level. We felt that this combination of sensors would be able to give someone substantial knowledge of what was happening in their home while they’re away.

We plan on allowing for other sensor systems to integrate with the Ubi, such as Sensordrone. The Ubi comes with RF chips that allow for communicating with these projects. Also, we’ve announced cooperation with SmartThings, which will make it easier to use the Ubi to read back information from sensors as well as drive home automation devices.
–
Ubi and SmartTings video
–

A: We want the Ubi to require as little intervention with hands as possible. However, some things are just easier to do with hands and eyes, such as inputting passwords or changing settings. To do these things, the Ubi will have apps for Android and iOS devices as well as a web portal for customizing the Ubi and making changes. It is possible to use voice alone to get the Ubi to access the Internet by spelling out the Wifi password. The real magic of the Ubi, however, will come when you can integrate your email or other platforms into it, and that’s easier to do online with a mouse and keyboard. Once setup, though, voice will be king and the Ubi will become that much more powerful with voice alone.
–

A: Absolutely. We feel that the Ubi is perfect for these tasks that would otherwise require you to constantly check your smartphone or PCs.
–

A: The Ubi is designed to be in a fixed location and to make that environment always on and ready, as well as easily monitored. We feel that on the go, current smartphones are more suited for accessing the Internet and communicating. Also, our phones are just for us whereas an Ubi could be used by anyone in that environment. However, we want to forget about smartphones once we step inside our homes. Maybe soon our smartphones could also disappear and there would be a virtual bubble around us that we could use through voice. Practically, when we’re walking around, talking into a computer is a little impractical, even though it’s available. Maybe something like an app for Google Glass?
–

A: There will be an app available for iOS and Android to set up and configure the Ubi. That should be released around the same time as the Ubi.
–

A: We’re currently working on being able to identify the voice of the speaker. With this, we can allow access to email or other personal settings on the Ubi.