It's a mess, we know

I apologize for our blog looking like a complete disaster at the moment and for most of the day. Unfortunately, some of the images used in the blog's theme are stored in Photobucket and as you can tell from the numerous messages throughout the site, we have exceeded our allocated bandwidth limit for the month. We are trying to correct the problem so please forgive us for the mess.

Update: Everything should look better now but the graphics are still missing. It will take a couple more hours before our new server is setup and able to stream the graphics at which point everything should be back to normal.

Update 2: Everything should be fine once again.

Robovie assistant robot

Japan's ATR has been working on an application for the Robovie robot platform that would see it perform assistant duties in shopping centers for lost patrons. The robot relies on a distributed sensor network consisting of cameras and time-of-flight sensors to monitor the behavior of up to 20 people at a time. If it determines that a particular person is lost then Robovie navigates to his/her location to offer assistance.

This is an interesting update on the museum tour guide robots that have been under development for nearly 10 years now; many of these robots such as CMU's Minerva have been successfully deployed in museums and remained operational over long periods of time. Something new that Robovie brings to the table is the use of an external sensor network for monitoring simultaneously a large number of people covering a considerably sized area. The downside of this approach is the need to introduce in the environment and calibrate a large number of sensors as opposed to relying only on sensors on-board the robot.

It will be interesting to see if Robovie will be a success and whether others will follow this approach when it comes to deploying autonomous assistant robots in large, crowded areas.

[Via Boing Boing]

Back after a brief absence

Regular readers probably noticed that I did not post any new stories this past week. The reason is that I was out of town enjoying some very much needed time off. I'm back now and I am looking forward to writing about the newest happenings in the world of robotics and artificial intelligence. Stay tuned!

COACH hand washing system: AI helping people with dementia

An AI system that we discussed last year has been named one of Advanced Imaging Magazine's Solutions of the Year. The computer vision and AI system, COACH, has been designed to aid sufferers of dementia complete the task of washing their hands while minimizing (and in some cases complete removing) the need for supervision by a caretaker. The system has undergone successful field trials since the last time we wrote about it (see our article "An artificial intelligence system for helping people with dementia" which includes a video of the system in use.) The team that consists of researchers in Canada and Scotland is confident that this technology will help those with cognitive disabilities live independently for longer increasing their quality of life. You can read a detailed description of the system components and the results of the field trials in Advanced Imagine Magazine's article Real-Time Audio-Visual Assistance for Dementia Sufferers.

NASA releases EUROPA under open source license

NASA's Ames Research Center (ARC) has released their planning software EUROPA (Extendable Uniform Remote Operations Planning Architecture) under an open source license welcoming the robotics community to use it and extend it.

EUROPA (Extendable Uniform Remote Operations Planning Architecture) is an advanced constrained based temporal planning platform [4]. Its architecture contains modules that can be used to solve problems in constraint satisfaction, planning, scheduling and combinatorial optimization.

EUROPA has been under development for years and NASA has successfully used it for a number of space missions including Deep Space One, Mars Exploration Rovers, and the International Space Station. Let me remind you that last year, NASA's JPL also released their CLARAty robot control architecture under an open source license.

You can find more information and download EUROPA at the official website here.

VideoTrace: Creating realistic 3D models from video

VideoTrace is a computer vision application that allows anyone to create detailed 3D models of complex objects from video using minimal user input. The work was presented in a paper last summer during SigGraph. The researchers use computer vision techniques known as structure from motion along with input from a user to extract 3D models of objects in video.

The user interacts with VideoTrace by tracing the shape of the object to be modeled over one or more frames of the video. By interpreting the sketch drawn by the user in light of 3D information obtained from computer vision techniques, a small number of simple 2D interactions can be used to generate a realistic 3D model. Each of the sketching operations in VideoTrace provides an intuitive and powerful means of modeling shape from video, and executes quickly enough to be used interactively. Immediate feedback allows the user to model rapidly those parts of the scene which are of interest and to the level of detail required.

The video below more than anything shows how easy to use and powerful VideoTrace truly is.


This interactive application can find use in many areas including creating 3D models for simulation, augmented reality, urban planning and others. VideoTrace is the result of a large collaborative effort among several schools in Australia. You can find more information at Australian Centre for Visual Technologies VideoTrace website here.

Ciao Robot documentary

The past year was the first time that much interest was shown globally on the issues of robots living with humans an event predicted to occur in the next 15-20 years. Roboethics is a fast growing field with a focus on studying the ethical, economic and social aspects of the age of robots. To create awareness of the issues before it is too late, Ciao Robot is a short documentary currently under production and featuring a number of prominent scientists discussing the relevant issues. You can watch the trailer below.

More information at the official Ciao Robot - The birth of Roboethics website here (unfortunately in Italian only.)

The beer brewing Bender replica

As I have mentioned before, I am a huge fan of the TV show Futurama (I loved the movie, btw) and so I could not pass posting about Simon Jansen's project to build a life-size replica of Bender. And this is not any version of Bender but rather the beer brewing one. In one of Futurama's episodes titled “The root of all evil,” the Planet Express crew brews beer in Bender's belly. Simon designed, built and brewed beer in a life-sized replica of the cartoon robot. He started the project in late June 2007, and he recently brewed his first batch of beer. The replica is not an autonomous device although it apparently has a computer brain and it can utter a few sentences. Regardless, this is a very cool project and a job well done. Simon has documented the construction of his Bender replica step by step and you can read it all here.

The photo of the Bender beer brewing replica robot is copyright Simon Jansen.

WowWee Tribot video

A video from Engadget showing WowWee's new Tribot robot toy. Apparently, it is capable of autonomous navigation (probably just collision avoidance though,) object recognition (I wonder if this has something to do with Evolution Robotics' object recognition technology,) and more than 10 minutes of audio for playback. It can also function in guard mode for guarding your house but this is most likely just for fun. And it will only cost $99. Watch the WowWee Tribot video below,

WowWee Rovio, Tribot, FemiSapien, Bug-bots and more

WowWee is at it again. They have just announced more than 20 new robot products at the Consumer Electronics Show (CES.)

Rovio is a remote controlled telepresence robot accessible using a web browser. It keeps localized using NorthStar which is probably the result of their partnership with Evolution Robotics last year. Nobody has been successful in marketing such a robot in the past and it will be interesting to watch what comes out of WowWee's efforts.

The new generation Robosapien will have wheels instead of legs and should be a good upgrade for the very popular line of robots. They will also make a female version of Robosapien,

The FemiSapien robot, a sophisticated humanoid interactive friend with distinctly fluid movements, features elegant gestures and dance moves controlled only by her 4-way joystick hands. She also features infrared, audio, and tilt sensors that enable her to synchronize her dance moves to music. The FemiSapien robot even speaks her own language of emotive sounds -- and responds to users' gestures, touch and sound.

WowWee is also venturing into companion robot territory with the introduction of Mr. Personality. The Mr. Personality robot should be lots of fun to play with an LCD screen displaying a very expressive face and additional personalities free to download from the web.


WowWee also introduced a new line of indoor flying robots to follow up on the success of the FlyTech Dragonfly. Many of these new flying robots are specifically targeted to young girls; they must think that this is a market worth pursuing. The flying robots will apparently have on-board logic and sensors allowing them to fly on their own without smashing into furniture and walls. It should be interesting to see how well this works because it is not an easy task. WowWee is also going to start selling a new series of bug-sized robots, the Bug-bots. These will be fun for experimenting with swarm formations.

Another cool product is the new universal remote control, RoboRemote, which as the Press Release says could be used with any of WowWee's robots. Best of all, these robots will be very affordable. Prices will range from $50 to $250.

There are too many interesting new robots that WowWee announced will be available in 2008. You can read the Press Release for more details. Also, PC World has an article on this.

Online machine learning repository

The UCI team from the Center for Machine Learning and Intelligent Systems has performed a major update and relaunch of the machine learning repository website; this is a great resource for those specializing in machine learning. The site hosts over 160 data sets for machine learning problems donated by research from around the world.

We have developed a new search portal that allows for faceted browsing of over 160 different data sets, using criteria such as data type, attribute type, number of instances, and number of attributes. The web site allows users to easily find data sets that satisfy particular criteria, such as "show all data sets containing only real-valued attributes and rank by number of instances." In collaboration with Rexa.info, the site now also provides links to published papers that have cited each data set.

We are always interested in new data set donations. If you have a data set that you would like to make available to the entire machine learning community, please feel free to donate using our automated web form. We thank all the previous donors for making this Repository a success.

Arthur Asuncion and David Newman, University of California, Irvine.

Visit (and bookmark) the UCI Machine Learning Repository.

Yuki-taro autonomous snowplow

Yuki-taro is another cool robot developed in Japan. This robot is designed to operate as an autonomous snowplow. It takes snow in the front and packs it into ice cubes. The robot relies on GPS and vision to navigate autonomously although I am not certain if it is already in use or just a prototype (probably the latter.) This is one robot that I can get a bit excited about just because it is actually designed to do something useful and also because given current technology I believe that it is possible to build. There is some more technical information available at the project website linked here translated from Japanese using the Google translation tool.

Our future

First of all, happy new year to all of our readers. I hope 2008 brings you and your family all the happiness in the world.

For the first post of 2008, I thought that there is nothing more appropriate than a lecture by futurist Ray Kurzweil talking about what he thinks our future will be like; and he is not making prediction about the very distant future but rather 10-15 years from. Granted the lecture in the video below is from 2005 but I still think it is very relevant and the best I could find on the subject for today's post. This is a summary of what his talk covers according to the informatiom on the TED website,

Prolific inventor and outrageous visionary Ray Kurzweil explains in abundant, grounded detail why -- by the 2020s -- we will have reverse-engineered the human brain, and nanobots will be operating your consciousness. Kurzweil draws on years of research to show the speed at which technology is evolving, and projects forward into an almost unthinkable future to outline the ways we'll use technology to augment our own capabilities, forever blurring the lines between human and machine.

And below is the video ( 23 minutes long and worth watching,)