Moving

right along. Next we are going to hear from Paolo Pirjaniand.


Paolo, no. He is the chief scientist and general manager of the robotics and vision group at Evolution Robotics and
he's going to talk about the evolving robotics. Thank you.

Thank you.

So we are actually a robotics company

but we are developing technologies that can be applied to other fields. So in a sense you could start by
defining what is robotics. And our view of robotics is not necessarily in line with the Hollywood portrayal
of R2D2 and so on. We believe that robotics is going to be mostly successfully embedded forms
such as you have with the microchips. You have microchips embedded into cell phones, pdas
and watches and so on and these form factors is not the main embodiment of
that technology and in the same way, we believe the same goes for robotics. So as softec technologists,
we develop these to enable machines to be able to see and move and so enabling the robots and otter machines to be able to
interact with a real environment using senses similar to vision.

And I guess, I don't have a presentation on this so I will just without the presentation. Let me tie this to
the previous talk which relates to search. The reason that I'm here today is because we were nominated for
one of our technologies, Vision technology, that was originally developed for Robotics but now
finding its way onto the cell phones for mobile visual search.

So we want to expand the search engines away from just using text as the main medium or way of
searching but using a much richer medium which is the pictures, the images. And following the
saying if a picture is worth more than a thousand words. And in many cases it is true.

So basically, the technology we are offering is called Viper for short. It stands for visual panoramic recognition.
And what you do is you take a picture and comparing it to a database of hundreds of
thousands of images and find a match. So it's visual search and you know there's
hundreds of millions of camera phones out there, so this is putting the world in your hand. And
the applications we are developing is basically allowing you to use this device to hyperlink the real world. So take a
picture of anything and that picture will be sent to a server. The server will do a visual pattern
matching and find similar or exact matches to scene or object you took a picture and download a response to
that. Sort of like typing in the name Paolo and finding a order list of matches. Here you
will take a picture of me or take a picture of a product or a scene or a monument and that image gets matched to the base
and downloads the information for you. So Softec can create the applications that we are deploying in the market with
partners today. I can demo this to you live afterwards to whoever is interested. Our for mobile marketing and for mobile
commerce applications. So imagine I go out to bookstore and I see this new book
on robotics and I'm interested in purchasing it but I want to read the reviews first though , I basically point my camera
phone to it, take a picture and immediately it will download a view of it or even link me directly to Amazon where I could do
price comparisons and say Amazon seems to be 5 dollars cheaper than Barns and Noble then inarucally purchase the book, it will
be sent to my home address that's tied into my subscription on my phone. Another thing would be, I see a new
advertisement poster of a new CD from Christina Aguilera and I take a picture of that and it could
download the Itunes right on to my cell phone and I can actually select the subset of songs on that CD rather than
purchasing the entire Cd, I'm paying 99 cents per song and I can select the ones I want and I have it right on my rocker
from Motorola and I can listen to it on my phone. And that those are the applications we are
deploying now. The business model we have is that we are primarily a technology company and we innovate in the areas of
computer vision and navigation technologies. We licensed these technologies to companies to develop the end
application or the end products. We have two partners in Asia that are deploying this kind of service in Asia. One is
in Japan and one in Korea and in next quarter Q1, they are going to announce
this applications and after they're deployed them, we have two partners in the U.S., one of them has
actually has already deployed this and next month you will see a press release and the application they are deploying is hosting
marketing campaign services using this technology. They have done campaign with a Jane
magazine, it's a female fashion magazine, and another one, L Girl. You can take a picture of any page on that magazine
it will recognize it and they will send to your home address samples of the product, the perfume or
lipstick and so on that is featured on that. So this is for product placement marketing and even counting the
number of eyeballs you have on your products.

So in addition to that, Evolution Robotics develops navigation technologies. So one of the challenges of
making robots really smart and having them do things intelligently in our homes is allowing them to
navigate, to move around the environment intelligently. The robotic products, for instance, you see
in the consumer space today, are moving on a random path. Robotic vacuum cleaners. One of them we developed
for sharper image and the other one from IRobot here. These robots don't know where they are
because they don't have GPS capability in doors. So we have developed a technology called Norstar which is a
in door GPS technology for prices you could fit in to consumer product. The state of the art
of localization before this technology was about 4 or 5 thousand dollars of beat up materials to develop a localization solution.
Norstar in consumer volumes which means hundreds of thousands and up all of you can get that for less than
twenty dollars. So that's what evolution does and thanks a lot for your attention.