Robots are a thing of the future video

Students from Hogg middle school (not sure which one it is but I suspect this one,) created the following short video of how they see robots as being a major part of humanity's future. Interestingly, they stumbled upon the idea of assistive robotics which currently is a very active area of research. Enjoy the video!

Keepon and HOAP-3 robots win top prize at ICRA '08

Two Japanese-built robots have captured first place at the ICRA ’08 Human-Robot Interaction Challenge. Keepon, designed for the National Institute of Information and Communications Technology in Japan, and HOAP-3, created by Fujitsu, tied for top spot when the results were revealed last week. Second place went to MIT and Intuitive Automata’s Robot Weight Loss Coach, third place went to Kaspar from the University of Hertfordshore, UK, and fourth spot was captured by Robotvie created by the University of Osaka, Japan.

The International Conference of Robotics and Automation is held annually and allows the world’s leading experts on robotics and smart software to gather and exchange their latest creations and breakthroughs. The Human-Robot Interaction Challenge competition assembled the top robots that coaxed the best responses to their appearance from human beings. The Keepon robots, which are squat, bright yellow and have wide eyes, have been used to gauge the reactions of young children. Cameras built-in to the Keepon’s eyes record the toddler movements as they watch the robot and a microphone built into its nose records the words spoken to the mechanical object. Child psychologists watch and observe the results which provide insight into how young children react to social interaction. On the opposite side of the coin, HOAP-3 has a humanoid appearance, stands 60 cms tall and looks more like a traditional robot in its appearance. HOAP-3 was designed by Fujitsu to be a competitor to Honda’s ASIMO concept robot of the future.

Next year’s conference, which will be held in Kobe, Japan, will have the theme of “Robotics and IRT for Livable Societies”.

Monkeys learn to feed themselves with robot arm

Two macaque monkeys that received a cutting edge implant in their brains have successfully learned how to manipulate a robotic arm to feed themselves. In a report published in “Nature” that describes the experiment, scientists working at the University of Pittsburgh and Carnegie Mellon University taught the pair of simians how to use a prosthetic arm using a joystick to feed themselves treats. The monkeys then had a medical procedure performed on them that placed a small grid-like structure (about the size of a freckle) over the motor cortex region of their brains that corresponds to arm movement. The grid’s sensors detected when this region of the monkeys brains would fire electrical impulses, meaning that the animals wanted to move their own arm. A computer would sense when the grid received stimulation from the monkeys’ motor cortex and then move the robotic arm. Over the course of several days after their operation the monkeys learned how to successfully operate the mechanical arm towards them, grasp food and then open the arm’s pincers. The animals had a success rate of about two-thirds of the time.

"In the real world things don't work as expected, the marshmallow sticks to your hand or the food slips, and you can't program a computer to anticipate all of that," said the senior author of the published report, Dr. Andrew Schwartz."But the monkeys' brains adjusted; they were licking the marshmallow off the prosthetic gripper, pushing food into their mouth, as if it were their own hand."

The implications of the experiment hold the potential for a day in the near future when humans without the use of their arms could be fitted with their own brain sensor that allowed them to operate robotic prosthetics. If humans can adapt as quickly as the monkeys did to learn how to use a robotic arm then the day may be not too far off when a similar technique is offered to paralyzed people.

NASA's Phoenix robot lands on Mars

After more than a few nail-biting moments, scientists at NASA rejoiced when the confirmation signal arrived from the Mars Phoenix Lander yesterday that it had successfully touched down on the red planet. After entering the planet’s atmosphere at 12,000 mph friction slowed the descent of the robot to one-tenth of that velocity before an automated parachute deployed. Seconds before planetfall a dozen rockets fastened to the underside of Phoenix fired to slow the craft down to a scant 5 mph, the craft’s final speed when it landed on Mars’ north polar region. Phoenix then transmitted its arrival back to worried scientists before beaming its first photos showing the Martian horizon and the deployment of its solar collectors.

Budgeted at $420 million dollars, the Phoenix will now begin its three-month long mission of exploration on Mars. Just as the Viking landers did back in the 1970s, the Phoenix will dig into the Martian soil and then perform a series of tests to detect if life ever existed on the fourth planet or if it may even still be there. In 2002 another robotic spacecraft, the Odyssey, discovered that a tremendous amount of frozen water was locked in the soil a few inches beneath the surface in Mars’ polar regions. Phoenix carried with it a oven in which Martian soil samples will be heated to a temperature of 1,200 degrees Fahrenheit. The scientists hope that the trace fumes analyzed by Phoenix will show if the ingredients for extraterrestrial life existed at one time on Mars.

Sending robotic spacecraft to Mars has been a risky proposition for the world’s governments. Nine years ago NASA met with failure when the Mars Polar Lander crashed while making a descent towards the planet’s south polar region. Phoenix is the first NASA Mars probe to use rocket thrusters to break its descent to the surface since the original Viking missions in 1976.

To see more photos that Phoenix has taken visit the NASA website located at http://phoenix.lpl.arizona.edu.

Open source Topological SLAM software available for download

Everyone knows that solving the Simultaneous Localization and Mapping Problem (SLAM) is essential to creating truly autonomous mobile robots. Many good SLAM solutions have been developed in the past 10 years but most of them rely on time-of-flight sensors (most often a SICK Laser) to achieve high localization accuracy and map resolutions. Visual SLAM approaches are also in development but currently a bit behind in terms of the size of the maps that can be successfully constructed.

The Oxford mobile robotics group has been a pioneer in monocular, real-time, visual SLAM solutions including metric and topological approaches. A couple of days ago, Dr Paul Newman and Mark Cummins released an open source version of their Fast Appearance Based Mapping (FABMAP) toolset. The C++ API is very easy to use. The results look impressive as you can see for yourself from the video below. If you are trying to enable your robot with SLAM abilities and know a bit of programming then you should give FABMAP a chance.

FabMap Algorithm - New College Dataset from Oxford Mobile Robotics Group on Vimeo.

Flame robot walks the walk

One of the biggest tricks in making productful human-like robots is understanding how to make them do one of the most basic physical motions of average human beings: walking. The engineers that made Honda’s Asimo robot spent the better part of two decades trying to imitate the motion of walking before producing the company’s famous goodwill mascot. Another fellow, Daan Hobbelen of the Dutch Delft University of Technology, has also come up with his own walking robot. Hobbelen has named his creation Flame. Standing 1.3 meters tall and weighing an compact 15 kilos, Hobbelen’s Flame robot may be the most intelligent walking robot in the world.

Flame employs several joints including dual sagittal knee, hip and ankle joints that use elastic parts to achieve torque. Sensors in Flame detect how well the robot is walking and then make incremental changes in the position of the robot’s body and legs to ensure that it doesn’t topple over. The brain of Flame is a Linux kernel running a control loop at a sample rate of 1 kHz. What Hobbelen calls Flame’s “organ of balance” helps the robot to gauge if it’s going to fall down as it moves forward; if so, Flame’s CPU will shift its feet, arms or torso to correct for the discrepancy. Right now the automaton has a speed of 0.45 m/s and can cross a floor with discrepancies of up to 8 mm. That may not win Flame any races with competing human walkers but it sets the stage for future improvements.

Hobbelen will earn his Ph.D from Delft in a ceremony scheduled for one week from today. Watch a video of Flame in motion below:

A robot that will climb the walls

Spider-Man has been doing it for decades and insects for millions of years. The ability to cling to walls has been a handy evolutionary feature for our buggy friends (and superheroes inspired by them) for a long time. Wouldn’t it be swell if we could invent our own machines that could climb walls without much effort? Well, evolution can move over because a team of researchers working at the non-profit outlet SRI International has designed an incredible wall-climbing robot that could have huge implications for future machines.

The SRI robot uses electro-adhesion to move up a ninety degree angle. Electro-adhesion is the process of generating a small electrostatic charge between the wall and (in this case) the robot’s body, allowing it to cling to surfaces. SRI will publicly unveil their creation this week at the International Conference on Robotics and Automation in Pasadena, CA but gave “Popular Mechanics” an exclusive first look at the device. The potential applications for wall-climbing robots are through the roof: security, home maintenance, the military, even kids toys. Think of what you would do if you were a five-year-old boy that wanted to freak out your little sister by having your spider robot climb up the wall behind her while she was reading.

According to an SRI researcher speaking to the magazine, the robot uses “a very small amount of power ... and shows the ability to repeatedly clamp to wall substrates that are heavily covered in dust or other debris.” Move over Roomba, this one doesn’t stop just at the walls.

MTMini: Make your own Surface computer

MultiTouch Mini (MTMini) is Seth Sandler's brainchild and you can think of it as make your own Microsoft Surface computer. Making MTMini, you can have tomorrow's multi-touch technology today and have all the fun of building it yourself for a cost of the price of buying a Surface computer. Seth has created MTMini as part of his undergraduate degree and his efforts to create a multi-touch musical interface.

This simple touch pad will cost you less than $50 for material and very little of your time. This is how Seth describes his engineering creation,

Don’t have a Multitouch Table yet? Have one, but need something smaller for testing? Building a small portable multitouch pad will allow you to test software and experiment on a smaller scale while building your full table or when away from your multi-touch screen. Have fun and make a MTmini! This uses Front Diffused Illumination, with normal ambient light (infrared not required or needed) and a normal off-the-shelf webcam (IR filter can still be in place).

The pad works by visually tracking the shadow the user's finger's cast on the screen. A simple idea that works well under mild lighting conditions. Seth has published all the details you need to build your own MTMini and he has even included the source you need. Obviously, the MTMini does not have the quality of the professionally designed Surface computer but with it you will have the satisfaction of having build it yourself.

The video below shows the MTMini in all its glory!

Birth simulator helps practicing doctors learn how to best deliver babies

Doctors in training are gaining valuable hands on experience with the process of childbirth thanks to a novel new robot that simulates childbirth. The BirthSIM was designed by scientists working in France and Canada so that it would simulate the birth of a child through a mother’s pelvis.

In some natural births the mother needs help in delivering her child and may require the doctors to use forceps or suction devices. The BirthSIM gives medical students an opportunity to learn how much force is needed to help assist with the birth of the baby. Electromagnetic sensors positioned in the head of the “baby” and in the forceps provide data to the operator, giving a 3D model of the inside of the birth canal and the relation of the forceps to the child’s head. A pneumatic arm, which pushes the mock baby’s head through the pelvis, can be tweaked to make the birth simulation less or more difficult for the medical student.

"We have been using a prototype to train obstetricians at the hospital in Lyon for a few years now," commented Richard Moreau, the head of the research team at Laboratoire Ampère in Lyon, France. "During three 1-hour sessions students use the forceps as often as they would during a normal three-year obstetric residency."

The students can also watch the 3D display to gain valuable insight when a skilled doctor performs the forceps delivery. By watching the more experienced doctor the students gain an understanding about how much force should be used.

Future iterations of the BirthSIM may incorporate changes into the simulated baby from the action of the forceps.

neuroArm neurosurgery robot makes medical history

The neuroArm medical robot made it into history's books after successfully assisting doctors perform delicate brain surgery to remove an egg-sized cancer tumour. This magnificent medical robot was developed at the University of Calgary Faculty of Medicine and it was in Calgary where Dr. Garnette Sutherland used it to operate on 21-year old Paige Nickason who after the successful operation is happily recovering at home. So, how does the neuroArm work?

neuroArm is the world’s first MRI-compatible surgical robot capable of both microsurgery and image guided biopsy. The surgical robotic system is controlled by a surgeon from a computer workstation, working in conjunction with intraoperative MR (magnetic resonance) imaging. Dr. Sutherland developed the intraoperative MRI machine with Winnipeg-based IMRIS Inc. The technology allows a high field MRI scanner to move into the operating room on demand, providing imaging during the surgical procedure without compromising patient safety. (source)

With a robot like neuroArm to assist them, doctors can perform complex surgery much easier with higher chances of success while patients suffer less from the operation and recuperate sooner. Another successful medical robot is the Da Vinci Surgical System designed by Intuitive Surgical. This is only the beginning. Expect to see the number of robots in surgery increase steadily over the next few years as technology improves and doctors (and patients) become more accustomed to the idea.

Survivor Buddy: An emotional rescue robot

Large scale disasters whether man-made or natural are a fact of life. Rescuers always have a difficult time locating and extracting survivors (the recent situations in earthquake ravaged China and hurricane stricken Burma are prime examples of this.) Robots offer much promise in making the job of rescuers easier helping save the lives of many people. Professor Robin Murphy from the University of South Florida has spent much of her time trying to develop effective rescue robots. Her robots have navigated the ruins of the World Trade Center in New York city after the 9/11 attacks and the darkest depths of the Crandall Canyon mine in Utah last year trying to help rescuers locate six trapped mine workers. While most rescue robotics research is focused on the development of semi-autonomous robots that can reach deep into disaster zones where human and animal workers cannot go in an effort to locate survivors, Murphy's new robot called the Survivor Buddy will take a different approach.

Survivor Buddy will be an emotional robot designed to keep victims company while rescuers try to reach them. Robots have been shown to be good companions for the elderly and Murphy believes that can do the same for disaster victims. More specifically (source,)

The Survivor Buddy would act as an emergency companion to people stuck in the crossfire of snipers or under the rubble of an earthquake-ravaged building like the ones now littering China.

She envisions a robot that plays soothing music to trapped victims and features a monitor showing the faces of loved ones and rescuers trying to reach them. It will deliver water and transmit a victim's vital signs to doctors. And it should be friendly, she said.

Professor Murphy has received a $500K research grant from Microsoft to start work on developing the new emotional rescue robots. Survivor Buddy is one of 8 projects selected as part of Microsoft's "Robots Among Us" research award on social robots. Microsoft is trying to encourage researchers to use their Robotics Studio platform via the distribution of awards which is a nice hook considering the difficulty researchers have in obtaining the necessary monetary support for their research programs. Whether this is ethical or not is a story for another post but as long as something good comes out of it, for example, effective rescue robots, then I guess we can all look the other way.

Meet the first synthetic person of Second Life

The digital world of Second Life is among one of the most popular internet destinations. People from around the world can construct an avatar of whatever person they want, be they male, female or otherwise, and then roam the landscape and interact with other explorers. Tens of thousands of Second Life players are always on at any given time, but now there is a unique citizen among the game’s population. His name is Edd Hifeng and the way he looks in Second Life, as a gray robot, isn’t too far off from what kind of intellect is behind the character. It turns out that the decisions and speech delivered by Edd is the product of an artificial intelligence program designed by researchers at the Rensselaer Polytechnic Institute in Troy, New York.

Edd is placed by his creators at different spots inside Second Life and the results of his interactions with other avatars controlled by real people are watched. While the programming that went into Edd’s thought processes is complex, he is able to function and respond to his environment at about the level of a four-year-old child. The students and teachers at Rensselaer are trying to push the envelope about what the limits are for designing a “smart” piece of software. While there is still a lot of work left to go to get to the point when a synthetic mind can be created that’s smart enough to interact with real humans on a variety of subjects, the Rensselaer team is hoping that their work on pushing Edd’s abilities even further and giving him emotions like courage or guile.

On their part the owners of Second Life don’t see any problem with Edd being one of their citizens and foresee the day when there are others like Edd living in cyberspace. “I think the real future for this is when people take these AI-controlled avatars and let them free in 'Second Life,'" commented John Lester, an operations manager for the company that owns Second Life. "Let them randomly walk the grid.''

Google Streetviews to start blurring faces

Google will finally address a major privacy issue inherent to their Stretviews product. Many people just don't like the idea that a snapshot of their daily activities can be made public on the World Wide Web. Some privacy groups have also been vocal about Google's silence on the privacy issues. The company has finally announced that they are taking steps to resolve these privacy issues by deploying face detection software to blur faces captured in the images.

I don't understand why it has taken Google so long to move in this direction. Robust face detection algorithms have been around for a while now and even though none has been deployed in a system as large as Streetviews, an early deployment would have saved the company much grief. Not to go off on a tangent, but an AP article on this story refers to this software as face recognition but this is probably a misuse of the term; face recognition implies recognizing individual people and such a solution would not be suitable for this application (not to mention that face recognition is a much harder problem than face detection to solve.)

Google will start rolling out the software in the next few months.

Read more:

• Where do Google Streetviews come from?


• Google Streetviews illegal in Canada?

Polar Robotic System up for MacRobert Award

It has been designed to safeguard millions of biological samples in sub-zero conditions and it’s also one of the finalists for this year’s MacRobert Award for technological and engineering accomplishments. Designed by the Automation Partnership, the Polar robotic system is a network of robotic armature apparatus and temperature components that stores and retrieves up to 10 million samples of blood, urine and other perishable medical items. The system works in temperatures of -80C, whether they are cooled by liquid nitrogen or standard refrigeration, and has been designed to last for a minimum of 25 years. By using a robot to do the work of sample retrieval, humans won’t have to suit up and enter the extreme cold environment and work time will be quickened. The machinery is already installed and working at the UK Biobank where cutting edge research into the prevention, diagnosis and treatment of diseases is underway.

The Polar System is competing for the MacRobert Award with three other innovations: a super-sensitive silicon sensor designed by Owlstone Ltd.; an advanced filter designed to remove soot from diesel engines; and Touch Bionics for the world’s first commercially available bionic hand, the i-LIMB. The prize is awarded by The Royal Academy of Engineering and the winner will receive a bronze medal as well as £50,000 in winnings. The winner of the 2008 MacRobert Award will be announced at the Academy’s Annual Awards Dinner on June 9, 2008.

Honda's ASIMO Robot and the Detroit Symphony Orchestra

In hopes to generate publicity for themselves and a good cause, Honda's ASIMO Robot conducted the Detroit Symphony Orchestra for a few minutes during a sold-out concert featuring cellist Yo-Yo Ma. Honda certainly know how to promote their technology. In my opinion, if they really want to impress people (and by that, I mean myself,) they should demo ASIMO organizing the books, papers, and post-it notes spread "neatly" all over my desk! At any rate, like I said earlier this was also about promoting DSO's efforts to help Detroit youth via their music programs so give credit to Honda for doing something nice for Detroit's young.

In conclusion, enjoy the video of ASIMO and the Detroit Symphony Orchestra making beautiful music together.

Additional information: Honda's ASIMO Robot Conducted Detroit Symphony Orchestra in "Impossible Dream"

Microtransat Challenge robotic fleet sets sail this October

This coming October there will be a novel nautical race taking place in the waters of the Atlantic Ocean. Eight robotic sailing ships will take part in the Microtransat Challenge, a three-month journey that will push the designers of these vessels to the limits of what robotics and artificial intelligence can presently accomplish.

The race begins from Viana do Castelo, Portugal and ends when the first boat crosses an imaginary finish line that stretches between St. Lucia and Martinique in the Carribean Ocean. Six countries are taking part in the challenge and among them is the ‘Pinta’ designed by Aberystwyth University in Wales. This 150-kilogram maritime robotic vessel will attain a maximum speed of about 7.4 kilometers per hour, or 4.6 mph. Solar panels will be used to gather energy for moving the ship’s tiller and pulleys while wind power will propel the boat across the big blue.

“This is the first time anybody has attempted to sail across any ocean with an automated boat. The big issue in robotics at the moment is longevity and flexibility in a complicated environment,” said Mark Neal of Aberystwyth University to The Times newspaper.” “Something that can survive for two to three months completely unassisted while doing something interesting is a major challenge. If it does get there I will be seriously cheerful. It will open up all the oceans to environmental monitoring by robots.”

Neal is talking about the possible future applications of smart ocean-travelling robotic craft which could be used to monitor changes in the ocean currents, acidity, pollution, carbon dioxide and chlorophyl content in the water and other meteorological data. Best of all is that since the ship would be unmanned there would be no danger to human life if weather conditions became dangerous on the high seas.

Robot performs complicated surgery techniques in Duke University tests

Researchers working at Duke University have reached a small but important milestone with the results generated from a tabletop robot ordered to perform a mock surgery. The machine was given the task to touch a needle mounted on the end of a robotic arm to the tip of a second needle placed inside a synthetic blood vessel graft. The software guiding the robot used a 3-D ultrasound transducer to locate the position of the needle that it needed to find. In short, the robot was smart enough to direct its focus to the correct place from the information being fed to it by the catheter transducer. This kind of delicate operation is routinely called upon from doctors in surgery and indicates that the latest advancements in collaborative A.I. software and robotics holds great potential for future medical science.

“In a number of tasks, the computer was able to direct the robot's actions,” said Stephen Smith, director of the Duke University Ultrasound Transducer Group and senior member of the research team. “We believe that this is the first proof-of-concept for this approach. Given that we achieved these early results with a rudimentary robot and a basic artificial intelligence program, the technology will advance to the point where robots – without the guidance of the doctor – can someday operate on people.”

Right now cardiologists that perform catheter-based surgeries need to use fluoroscopy to find the correct locations within the patient. This kind of medicine requires that a radioactive dye be injected into the patient to gain the data needed for the surgery, but with a robot that is able to process the data from a transducer and act on it in real-life while performing the surgery, the surgical technique could be made faster and less complicated. “Putting a 3-D ultrasound transducer on the end of the catheter could provide clearer images to the physician and greatly reduce the need for patients to be exposed to radiation,” offered Smith.

But there is still much research to be done before robots are working alongside doctors in the O.R. The next step for the Duke University team is to take what they’ve learned and apply it for animal trials. The group’s research now appears in the IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency Control journal which is available online.

Everything about robot sex that you want to know but have been afraid to ask

As smart machines become increasingly smarter and the continual evolution of robotics produces more lifelike mechanical creations, when will the time come when humans begin to have sexual relationships with robots? It may seem like a topic best left for the imagination of a fetish mag right now but to Dr. David Levy, it’s his business to think about the implications. Levy has just written a new book titled “Love and Sex with Robots: The Evolution of Human-Robot Relationships” where the artificial intelligence expert tackles the controversial and dicey prospect of mechanical-biologicial hanky panky. Levy firmly believes that as the technological difficulties of robot engineering and creating artificial emotions begins to be solved humans will begin to have physical relationships with these sexbots. There may even come a day when the debate over marriage should be extended to cover artificial personalities.

In an interview with the relationship website Lavalife, Levy postulates that the day will come when society considers sex with robots as normal as checking one’s email. “If a robot behaves as though it has consciousness, and is convincing in this behavior, then people will accept it as having consciousness. That will make the robot more appealing as a friend or lover,” explains the doctor. It’s already been shown that people will mentally view a robot that displays human-like characteristics as having its own personality, but in the future, when computing power is many times what its limit is today, robots will be able to fully mimic the behavior of a sentient being. That means for those that want to have their ideal sex partner, someone that will fulfill their every fantasy without hesitation and with enthusiasm, the lure of a robot relationship may prove to be irresistible. But won’t the illusion be wrecked if the participant knows in the back of their mind that it’s intimacy with a machine? “At that point it will not matter that the emotions are artificial,” Levy explains. “What will matter is what the human user experiences.”

A sex robot’s personality could even be downloaded into a new body at a later time. Levy has even imagined the day when the original owner of the robot passes away and that its future partner/owner purchases it because “Your robot can then be chosen by another human being who specifically wants a robot that has already experienced living with you and has learnt from that experience.”

Robotic radiosurgery course doubles enrollment

As robotics becomes more refined so too are the advancements in medicine. The organizers of the Robotic Radiosurgery Course held on April 11 and 12 in Boston, Massachusetts, are seeing this first hand as the number of students attending this year’s course nearly doubled in size versus the turnout for the course held in 2007. More than 175 health care professionals from 13 different countries attended the two day course to learn about how they or their medical facility could use the CyberKnife radiosurgery robot to combat cancers and tumors in a non-surgical manner.

The CyberKnife stereotactic radiosurgery system uses a linear accelerator on a robotic arm which then delivers concentrated beams of radiation to very specific zones on the patient. This allows the doctor to attack tumors in regions of the patent’s body which were previously inaccessible, such as the spine or within the brain. Without the need for incisions there is no recuperative time required by the patient, and the highly focused beams of radiation are centered so they only impact the tumor and none of the surrounding tissue. The software used by the CyberKnife robot incorporates the data from image guidance technology which automatically keeps a track of not just the present location of the tumor as the arm moves around the patient’s body but the actual real-time movement of the patient and counters for this discrepancy.

According to the American Cancer Society's Cancer Facts & Figures for 2007, 1.4 million new cancer cases were diagnosed last year in the United States alone. Many of the people diagnosed with this dreaded disease can be treated by using radiotherapy. Evidence is showing that the hospitals that use the CyberKnife robotic technology can nearly double the amount of patients that can receive this sort of treatment. When added together with the benefits of being able to treat the tumor without even breaking the skin of the patient, the savings in post-recovery time, resources and manpower for the medical system are plain to see. There are now just under 100 facilities worldwide set up with the CyberKnife system and over 20,000 patients have been treated using this advanced technology.

Second Life enters real life with augmented reality project

Second Life is undoubtedly one of the most popular massively multiplayer online (MMO) worlds. People log-on to play the game often enough that some consider it an addiction no different than alcohol or drugs. Regardless, researchers from the Georgia Institute of Technology are working on bringing Second Life's virtual world to the real world via a new software client and augmented reality helmet. The project is called AR Second Life.

In this research project, we demonstrate that MMOs also provide a powerful platform for Augmented Reality (AR) applications. We introduce the notion of AR Stages, persistent, evolving spaces which encapsulate AR experiences in online three dimensional virtual worlds.

Based on the 3D virtual world Second Life, our custom client software blends together locations in physical space with corresponding places in the Second Life virtual space and leverages the power of the MMO to create an powerful AR authoring environment targeted at a wide audience and fields of application.

Users experience the blended virtual and real worlds using eMagin's Z800 head-mounted display (HMD) equipped with a small Point Grey Research camera (Firefly MV) which according to the company is used together with a tracking device used for estimating the helmet's position and orientation to make the AR Second Life experience possible.

The camera is mounted on the front of the HMD, pointing down and reflected forward using a right-angled prism to allow the camera to be optically closer to the person's eyes than would otherwise be possible. An IS-1200 hybrid tracking device from Intersense is used to track the exact position and orientation of the HMD precisely over a large area.

The following video describes the system in more detail including example outputs, i.e., example views from the user's point of view (if you ever wanted to see a Second Life avatar enter the real world then this is your chance.)



Expect augmented reality to be much more common in the not so distant future. As the technology improves both in software (better rendering and structure estimation) and hardware (cheaper, smaller, and faster hardware,) augmented reality systems similar to AR Second Life will become commonplace until the virtual worlds completely take over. Eventually, the hardware will be small enough to be embedded inside our bodies and output will be directly to our brain instead indirectly via a bulky HMD.

This is a future that has been predicted by many and it might scare some of us but it seems inevitable to me. In such virtual worlds much of the uncertainty in sensing the real world will be eliminated and AI algorithms will become applicable; inevitably AI will come to dominate the virtual worlds. Are we going to be trapped in them or continue our dominion over all other intelligences (real and artificial) on our planet? Now that I think about it, all these sound like elements from the movie The Matrix. Neo, please take the red pill!

ESA's MEXAR2 AI improves our exploration of Mars

Scientists at the European Space Agency are relishing in the commendations received over their new smart software being used to send back valuable data from the red planet. Back in January 2005 the ESA’s Mars Express slipped into orbit around Mars and began sending back data about the world’s dry, frozen landscape, its thin atmosphere and even readings about what may lie underneath the ground of the planet. The data was then transmitted back to Earth but had to be sent in a carefully controlled manner. If the sequence of transmission were out of order or human error came into the reception then the data would be forever lost, the reason being that the Mars Express orbiter only has a limited amount of memory storage capabilities since new data is constantly being written over the older data.

Until recently the only way for the data to be successfully received was by sending authorization to the spacecraft, a lengthy and elaborate series of commands that took up valuable manhours and was always on a different schedule as the two planets constantly moved through their orbits. Now the data retrieval method is much more easier thanks to a smart application called MEXAR2 which stands for Mars Express AI Tool. Developed by the Institute for Cognitive Science and Technology in Italy, the solution reduces the manpower required to complete the data dumps by half from the old method. It also frees up the computer and bandwidth time needed to communicate with the spacecraft and has automated some of the command decisions that were required from human operators – and that means no more loss of data. "With MEXAR2, any loss of stored data packets has been largely eliminated," said Fred Jansen, the mission manager for Mars Express.

The ESA wants to continue using AI technology for its future robotic missions to Mars, such as its planetary rover mission, and it is also using the application for boosting productivity for the data it sends to the spacecraft.