Wednesday, December 08, 2004
Sony exoskeleton and Aichi EXPO 2005
Exoskeletons are becoming a very interesting side-branch of robots that jump. To the DARPA BLEEX project at UC Berkeley and the SERVO exoskeleton challenge we now must add Toyota's amazing iFoot exoskeleton. This system, detailed in a recent press release, is about 10 feet high. A seated individual uses a joystick to move around on huge legs, apparently allowing a limited form of dancing. Toyota introduced the iFoot along with a hybrid wheelchair/small car called the iUnit. In addition, Toyota expanded its line of partner robots to include a drum-playing, two-wheeled robot, a tuba-playing robot and a robot 'DJ' which interacts with the crowd during musical performances. This very impressive set of robotic devices put Toyota firmly in the lead compared to Honda, at least until Honda introduces its upgraded Asimo early in 2005.
June 9-19 2005 will be when you can see the Toyota robots, plus other Japanese robots - provided you can get a ticket to EXPO 2005 in Aichi, Japan. On a page entitled "We live in the robotic age", the festival website shows the layout of the Expo, in particular detailing the locations of working robots, including garbage collection robots, childcare robots, cleaning robots, tourguide robots, and security robots. In addition, prototype robots not ready for commercial work will also be demonstrated. One of the prototypes is a "dance partner robot" - if it works, this is certainly a robot that jumps!
Exoskeletons are becoming a very interesting side-branch of robots that jump. To the DARPA BLEEX project at UC Berkeley and the SERVO exoskeleton challenge we now must add Toyota's amazing iFoot exoskeleton. This system, detailed in a recent press release, is about 10 feet high. A seated individual uses a joystick to move around on huge legs, apparently allowing a limited form of dancing. Toyota introduced the iFoot along with a hybrid wheelchair/small car called the iUnit. In addition, Toyota expanded its line of partner robots to include a drum-playing, two-wheeled robot, a tuba-playing robot and a robot 'DJ' which interacts with the crowd during musical performances. This very impressive set of robotic devices put Toyota firmly in the lead compared to Honda, at least until Honda introduces its upgraded Asimo early in 2005.
June 9-19 2005 will be when you can see the Toyota robots, plus other Japanese robots - provided you can get a ticket to EXPO 2005 in Aichi, Japan. On a page entitled "We live in the robotic age", the festival website shows the layout of the Expo, in particular detailing the locations of working robots, including garbage collection robots, childcare robots, cleaning robots, tourguide robots, and security robots. In addition, prototype robots not ready for commercial work will also be demonstrated. One of the prototypes is a "dance partner robot" - if it works, this is certainly a robot that jumps!
Monday, December 06, 2004
Cybercars & why robotics will trump computer game animation
First off, an awesome report in Time Europe about the commercial debut of "cybercars" - autonomous, driverless cars that can navigate on normal city streets using lidar and other sensors. Compared to pie in the sky US ideas, the Cybercar (http://www.cybercars.org ) is already a practical product. According to the Time Europe article,
"...The CyberCar was developed by a consortium of 15 European research centers and spearheaded by Michel Parent, program director for research and development at France's INRIA, an automation and robotic research center...."
and,
"...The town of Antibes on the French Riviera and the nearby principality of Monaco are considering buying their own fleets to taxi visitors around their cramped streets...."
This is amazing work - even more exciting, in its way, than the DARPA Grand Challenge 2005. The consortium tested the Cybercars on ordinary people during June 2004, with favorable results. The cars show good obstacle avoidance, and even display co-operative behavior with each other. All this with Lidar - imagine what they'll be like with a low-cost computer vision system! It may be that Europe will join Japan and Korea as the main centers in the robotic revolution, while the US snags last place.
Which brings us to our next point. It turns out that 2005 will see the debut of two movies done entirely in computer animation a la Toy Story - Cars, and Robots (I'll leave it to you to find these movies' lame websites). These computer-animation films,created with programs like Maya, are the US answer to real-world robots - humanoid and automotive. While Japan steadily advances its humanoid robotic technology and the EU makes driverless cars a reality, the US continues to chase cyberspace. It is likely that many people will see movies like Robots when it comes out - without the slightest idea that real-world robots now exist!
Much has been said about the US exporting its "heavy" industries to countries like China while remaining tops in media production. The fact that there is no decent US counterpart to the EU Cybercars shows that this is true at a very deep level. Instead of real-world machines, the US is still focused on the fantasies of cyberspace, asset inflation (read:finance), to its long-term detriment.
I submit that, in the 5-10 year timeframe, people will lose interest in computer-generated characters for real robotic ones.
The reasons for this are not hard to understand. Programs like Maya do not allow realistic animation, despite all the blather among techno-gurus about "hyperreal games" and Final Fantasy like synthespians. In practice, the virtual worlds that CGI movies are crafted in utterly lack real-world physics. Movements in these fantasy worlds are not realistic at all, and could never happen in the real world. All the hoopla around Final Fantasy ignores this fact - the movie is a work of art, and not in any way "real" looking. The gadget-fondling tech gurus which seem to sprout everywhere these days somehow don't seem to notice.
In fact, CG programmers can't really do effective 3D motion at all. Trapped with a 2-dimensional mouse as their primary input device, CG programmers make due with a bag of automatic 3D movement scripts written into the programs. They don't animate the detailed motion of a CG creature - instead, they specify broad movements, and the program "tweens" the details. The details, computer-drawn are non-physical and have a very "puppetlike" appearance.
This use of common "motion" scripts causes all CG films to have a common "look and feel", which is very puppet-like. CG characters move more rapidly, non-physically, and at speeds impossible for real-world objects. True, this is art - and part of the charm of CG movies is the very limitations of the CG medium which imposes these non-realistic kinds of movements on CG characters. This is the charm of traditional animation as well. I don't want to knock CG as art, but simply point out that it is no more real than a 1930s Bugs Bunny cartoon.
But imagine it is 2010. Instead of virtual comptuer characters dancing on the other side of a computer monitor we have physical, real-world robots dancing on stage. Unlike CG characters, they have to cope with, and react to, real-world physics. Their movements are "realistic" - they have to cope with gravity in ways similar to living things. They don't have any of the fantasy charm of a CG character - instead, their movements create a haunting impression of a "ghost in the machine."
Now, consider an audience jaded by almost 20 years of CG animation. While some progress has been made, in fact there has been little real change in CG style since the classic Dragonheart movies of the mid-1990s. While technology may have changed, the artistic "look and feel" of CG animation has hardly budged. And after 20 years, an entire generation will have grown up with this particular "look and feel" - and be hungry for something different.
And along will come real-world robots, doing the same things CG characters do, but with a difference: they are real. Imagine the pull of entertainment robotics at this juncture - it will feel like, and in fact will be, a new artistic medium. Faster than might be anticipated, the new generation will abandon the increasingly stale, monoculture world of so-called "realistic" videogames for robots.
To coin a phrase for this next stage of pop culture revolution, we are about to take a "turn to the real". Not my idea, but originally experessed by Dan Danknick of Servo Magazine, the "BYTE" magazine of robotics.
After all, who would want to play with a virtual robot walled off behind the plastic of a plasma screen, when they can dance with a real one?
This will be a revolution. which those countries pushing robotics will win - they will help to define art and style in the 2010 - 2030 timeframe and not just technology. And, unless major changes occur by then, the US will not be one of the leaders.
First off, an awesome report in Time Europe about the commercial debut of "cybercars" - autonomous, driverless cars that can navigate on normal city streets using lidar and other sensors. Compared to pie in the sky US ideas, the Cybercar (http://www.cybercars.org ) is already a practical product. According to the Time Europe article,
"...The CyberCar was developed by a consortium of 15 European research centers and spearheaded by Michel Parent, program director for research and development at France's INRIA, an automation and robotic research center...."
and,
"...The town of Antibes on the French Riviera and the nearby principality of Monaco are considering buying their own fleets to taxi visitors around their cramped streets...."
This is amazing work - even more exciting, in its way, than the DARPA Grand Challenge 2005. The consortium tested the Cybercars on ordinary people during June 2004, with favorable results. The cars show good obstacle avoidance, and even display co-operative behavior with each other. All this with Lidar - imagine what they'll be like with a low-cost computer vision system! It may be that Europe will join Japan and Korea as the main centers in the robotic revolution, while the US snags last place.
Which brings us to our next point. It turns out that 2005 will see the debut of two movies done entirely in computer animation a la Toy Story - Cars, and Robots (I'll leave it to you to find these movies' lame websites). These computer-animation films,created with programs like Maya, are the US answer to real-world robots - humanoid and automotive. While Japan steadily advances its humanoid robotic technology and the EU makes driverless cars a reality, the US continues to chase cyberspace. It is likely that many people will see movies like Robots when it comes out - without the slightest idea that real-world robots now exist!
Much has been said about the US exporting its "heavy" industries to countries like China while remaining tops in media production. The fact that there is no decent US counterpart to the EU Cybercars shows that this is true at a very deep level. Instead of real-world machines, the US is still focused on the fantasies of cyberspace, asset inflation (read:finance), to its long-term detriment.
I submit that, in the 5-10 year timeframe, people will lose interest in computer-generated characters for real robotic ones.
The reasons for this are not hard to understand. Programs like Maya do not allow realistic animation, despite all the blather among techno-gurus about "hyperreal games" and Final Fantasy like synthespians. In practice, the virtual worlds that CGI movies are crafted in utterly lack real-world physics. Movements in these fantasy worlds are not realistic at all, and could never happen in the real world. All the hoopla around Final Fantasy ignores this fact - the movie is a work of art, and not in any way "real" looking. The gadget-fondling tech gurus which seem to sprout everywhere these days somehow don't seem to notice.
In fact, CG programmers can't really do effective 3D motion at all. Trapped with a 2-dimensional mouse as their primary input device, CG programmers make due with a bag of automatic 3D movement scripts written into the programs. They don't animate the detailed motion of a CG creature - instead, they specify broad movements, and the program "tweens" the details. The details, computer-drawn are non-physical and have a very "puppetlike" appearance.
This use of common "motion" scripts causes all CG films to have a common "look and feel", which is very puppet-like. CG characters move more rapidly, non-physically, and at speeds impossible for real-world objects. True, this is art - and part of the charm of CG movies is the very limitations of the CG medium which imposes these non-realistic kinds of movements on CG characters. This is the charm of traditional animation as well. I don't want to knock CG as art, but simply point out that it is no more real than a 1930s Bugs Bunny cartoon.
But imagine it is 2010. Instead of virtual comptuer characters dancing on the other side of a computer monitor we have physical, real-world robots dancing on stage. Unlike CG characters, they have to cope with, and react to, real-world physics. Their movements are "realistic" - they have to cope with gravity in ways similar to living things. They don't have any of the fantasy charm of a CG character - instead, their movements create a haunting impression of a "ghost in the machine."
Now, consider an audience jaded by almost 20 years of CG animation. While some progress has been made, in fact there has been little real change in CG style since the classic Dragonheart movies of the mid-1990s. While technology may have changed, the artistic "look and feel" of CG animation has hardly budged. And after 20 years, an entire generation will have grown up with this particular "look and feel" - and be hungry for something different.
And along will come real-world robots, doing the same things CG characters do, but with a difference: they are real. Imagine the pull of entertainment robotics at this juncture - it will feel like, and in fact will be, a new artistic medium. Faster than might be anticipated, the new generation will abandon the increasingly stale, monoculture world of so-called "realistic" videogames for robots.
To coin a phrase for this next stage of pop culture revolution, we are about to take a "turn to the real". Not my idea, but originally experessed by Dan Danknick of Servo Magazine, the "BYTE" magazine of robotics.
After all, who would want to play with a virtual robot walled off behind the plastic of a plasma screen, when they can dance with a real one?
This will be a revolution. which those countries pushing robotics will win - they will help to define art and style in the 2010 - 2030 timeframe and not just technology. And, unless major changes occur by then, the US will not be one of the leaders.
