ROBOT FEATURES

Seeing through concrete
Already in existence are sensors the size of matchboxes which respond to heat, light, movement or sound; and a variety of programmes, including one called Smart Dust, are working on further miniaturising these and improving their ability to work as networks. A dozen US university teams are also developing micro-aircraft, weighing a few grams each, that imitate birds and insects and could carry sensor equipment into specific buildings or rooms.
Darpa's VisiBuilding programme, meanwhile, is making "X-ray eye" sensors that can see through concrete, locating people and weapons inside buildings. And Human ID at a Distance is working on software that can identify individual people from scans of their faces, their manner of walking or even their smell, and then track them anywhere they go.
Closely related to this drive are projects involving compu-ter simulations of urban landscapes and entire cities, which will provide backdrops essential for using the data gathered by cameras and sensors. The biggest is Urban Resolve, a simulated war against a full-scale insurgency in the Indonesian capital, Jakarta, in the year 2015.
Digitised cities
Eight square miles of Jakarta have been digitised and simulated in three dimensions. That will not surprise computer gamers, but Urban Resolve goes much further: the detail extends to the interiors of 1.6 million buildings and even the cellars and sewers beneath, and it also includes no fewer than 109,000 moving vehicles and people. Even the daily rhythms of the city have been simulated. The roads, says one commentator, "are quiet at night, but during weekday rush hours they become clogged with traffic. People go to work, take lunch breaks and visit restaurants, banks and churches."
Digitise any target city and integrate this with the flow of data from many thousands of sensors and cameras, stationary and mobile, and you have something far more powerful than the regular snapshots today's satellites can deliver. You have continuous coverage, around corners and through walls. You would never, for example, lose those mortar bombers who got out of their car and ran away.
All this brings omniscience within reach. The US web-based magazine DefenseWatch, which monitors developments in strategy and hardware, recently imagined the near-future scenario of an operation in the developing world in which a cloud of minute, networked sensors is scattered like dust over a target city using powerful fans. Directed by the sensors, unmanned drones patrol the city, building up a visual and audio picture of every street and building. "Every hostile person has been identified and located," continues the scenario. "From this point on, nobody in the city moves without the full and complete knowledge of the mobile tactical centre."
Another Darpa project, Integrated Sensor is Structure, is working on the apex of such a system: huge, unmanned communications and surveillance airships that will loiter above target areas at an altitude of 70,000 feet - far above most airline traffic - providing continuous and detailed coverage over a whole city for a year or more.
From these platforms, all the information could be fed down in real time to soldiers and commanders carrying the hand-held computers being developed by the Northrop Grumman Corporation with Darpa funding. The real aim, however, is not to expose flesh-and-blood Americans on the ground, but where possible to use robots. That way there will be no "body bag problem"; and in any case machines are better equipped than human beings to process and make use of the vast quantities of data involved.
In one sense, robots are not new: already, armed drones such as Predator, "piloted" by CIA operators from screens in Florida, have been responsible for at least 80 assassination raids in Iraq, Afghanistan, Yemen and Pakistan (killing many civilians as well). Defence contractors have also developed ground-based vehicles capable of carrying cameras and weapons into the battlefield.
But this is only the start. What will make the next generation different is that they are being designed so that they can choose, all on their own, the targets they will attack. Operating in the air and on the ground, they are being equipped with Automated Target Recognition software capable not only of comparing signals received from new-generation sensors with databases of targets, but also of "deciding" to fire guns or launch missiles automatically once there is a good "fit". Automated killing of this kind hasn't been approved by anyone yet, but it is certainly being planned. John Tirpak, editor of Air Force Magazine in the US, expects initially that humans will retain the last word, but he predicts that once robots "establish a track record of reliability in finding the right targets and employing weapons properly", the "machines will be trusted to do even that".
Planners believe, moreover, that robot warriors have a doomsday power. Gordon Johnson, a team leader on Project Alpha, which is developing robots for the US army, predicts that, if the robot's gun can return fire automatically and instantly to within a metre of a location from which its sensors have detected a gunshot, it will always kill the person who has fired. "Anyone who would shoot at our forces would die," says Johnson. "Before he can drop that weapon and run, he's probably already dead. Well now, these cowards in Baghdad would have to pay with blood and guts every time they shoot at one of our folks. The costs of poker went up significantly. The enemy, are they going to give up blood and guts to kill machines? I'm guessing not."
Again, this may sound like the plot of a B-movie, but the US military press, not a body of people given to frivolity, has been writing about it for some time. DefenseWatch, for example, also featured robots in that future war scenario involving sensors dispersed by fans. Once a complete picture of the target city is built up, the scenario predicted, "unmanned air and ground vehicles can now be vectored directly to selected targets to take them out, one by one".

AMERICA"S ROBOT ARMY

Already there are killing machines operating by remote control. Soon the machines will be able to kill on their own initiative. A new warfare is on its way

War is about to change, in terrifying ways. America's next wars, the ones the Pentagon is now planning, will be nothing like the conflicts that have gone before them.

In just a few years, US forces will be able to deal out death, not at the squeeze of a trigger or even the push of a button, but with no human intervention whatsoever. Many fighting soldiers - those GIs in tin hats who are dying two a day in Iraq - will be replaced by machines backed up by surveillance technology so penetrating and pervasive that it is referred to as "military omniscience". Any Americans involved will be less likely to carry rifles than PlayStation-style consoles and monitors that display simulated streetscapes of the kind familiar to players of Grand Theft Auto - and they may be miles from where the killing takes place.

War will progressively cease to be the foggy, confusing, equalising business it has been for centuries, in which the risks are always high, everyone faces danger and suffers loss, and the few can humble the mighty. Instead, it will become remote, semi-automatic and all-knowing, entailing less and less risk to American lives and taking place largely out of the sight of news cameras. And the danger is close to home: the coming wars will be the "war on terror" by other names, conflicts that know no frontiers. The remote-controlled war coming tomorrow to Khartoum or Mogadishu, in other words, can happen soon afterwards, albeit in moderated form, in London or Lyons.

This is no geeky fantasy. Much of the hardware and software already exists and the race to produce the rest is on such a scale that US officials are calling it the "new Manhattan Project". Hundreds of research projects are under way at American universities and defence companies, backed by billions of dollars, and Donald Rumsfeld's department of defence is determined to deliver as soon as possible. The momentum is coming not only from the relentless humiliation of US forces at the hands of some determined insurgents on the streets of Baghdad, but also from a realisation in Washington that this is the shape of things to come. Future wars, they believe, will be fought in the dirty, mazy streets of big cities in the "global south", and if the US is to prevail it needs radically new strategies and equipment.

Only fragments of this story have so far appeared in the mainstream media, but enough information is available on the internet, from the comments of those in charge and in the specialist press to leave no room for doubt about how sweeping it is, how dangerous and how imminent.

Military omniscience is the starting point. Three months ago Tony Tether, director of the Defence Advanced Research Projects Agency (Darpa), the Pentagon's research arm, described to a US Senate committee the frustration felt by officers in Iraq after a mortar-bomb attack. A camera in a drone, or unmanned aircraft, spotted the attackers fleeing and helped direct US helicopters to the scene to destroy their car - but not before some of those inside had got out. "We had to decide whether to follow those individuals or the car," he said, "because we simply didn't have enough coverage available." So some of the insurgents escaped. Tether drew this moral: "We need a network, or web, of sensors to better map a city and the activities in it, including inside buildings, to sort adversaries and their equipment from civilians and their equipment, including in crowds, and to spot snipers, suicide bombers or IEDs [improvised explosive devices] . . . This is not just a matter of more and better sensors, but, just as important, the systems needed to make actionable intelligence out of all the data."

Darpa has a host of projects working to meet those needs, often in surprising ways. One, called Combat Zones That See, aims to scatter across cities thousands of tiny CCTV cameras, each equipped with wireless communication software that will make it possible to link their data and track the movements of every vehicle on the streets. The cameras themselves will not be that different from those found in modern mobile phones

ROBOTS CONTINUE

Robots and Positioning

The Waseda Humanoid Robot Project has developed different types of robots. For human-robot symbiosis, we consider the physical function of the robots, including safety control and human-robot contact identification based on tactile recognition. Equally important to consider is their mind acquisition, such as intelligence and emotion. The developed robots are categorized as human-assisting humanoid robots, biped humanoid robots, intelligent robots, and emotional communication robots.

Japan's Graying Population

The team carefully studied the robot's symbiosis with the natural environment. Some robots are intended for autonomous movement and have self-support of energy so that they can survive in the natural environment and work closely with humans.

One of the main functions for most of these robots is the ability to define its position and to navigate. To provide a robot with the means to navigate is a huge undertaking — the team is working on only the first of many steps to solve this problem. This first step is to use various sensors — inertial sensors, satellite navigation system receivers, magnetic sensors and others — operating in coordinate domain. These sensors provide the robot with coordinates and coordinate-related parameters such as azimuth and orientation

HUMANOID ROBOT

Waseda University established the Humanoid Robotics Institute in April 2000 to promote research that aims to construct a new relationship between humans and machines in an advanced information society. As researchers, we expect that in sometime in this century robots will provide housework assistance for the elderly, a well as entertainment and other functions to improve the quality of life for humans. To make possible this symbiosis between humankind, robot, and the environment, we need to build accommodations into a house's structure and functions.


THE WASEDA: Humanoid Robot Project has developed different types of robots for different needs. Models include (from left) WABOT-1 from 1978, Wendy from 1999, and WABOT-2 from 1984. Hadaly-2 from 1997 is pictured at left.

The Waseda University WABOT-HOUSE (WAseda roBOT HOUSE) project, directed by Shigeki Sugano, was established to facilitate this research. The special test facilities for this project were created in Gifu Prefecture. In 2001 Waseda University established the WABOT-HOUSE Laboratory, located in Techno Plaza R&D site in Kagamigahara City, in the middle of an industrial region. Gifu Prefecture initiated the project with the expectation that the robotics industry can provide a significant economic boost to the region. The Japanese government took the step to designate Kagamigahara City a special robotics zone.

As a part of this project, the Waseda University team developed not only a robotic component, but also design theories of environment space, construction, and social systems for human-robot symbiosis. These theories were implemented in the structures and functions of houses and facilities to allow an integration of robot technology with various environments. Fine artists joined with robotics engineers, architects, and IT researchers to take part in the WABOT-HOUSE project. Together, these experts are trying to design an optimal system from both robotic and architectural points of view to realize truly practical symbiosis between human and robots.

As part of a system to enable the robots to navigate their environment, a GPS and pseudolite solution combined with other technologies is being investigated

ROBOT LIFE

For more than 30 years a dedicated team of scientists and engineers at Waseda University in Tokyo, Japan, have been working on a project that will integrate robots into our everyday lives. One of the main functions for most of these robots is an ability to define its position and navigate.




The first step is to use different sensors, such as inertial sensors, satellite navigation system receivers, magnetic sensors, RFID tags, and others, which are operating in coordinate domain. These sensors provide robot with coordinates, and coordinate-related parameters such as azimuth and orientation. The navigation process using these sensors can be in general terms described as an artificial navigation.

The Waseda Humanoid Robot Project, headed by Professor Shuji Hashimoto, is researching the integration of robots into our social infrastructure, or human-robot symbiosis. The advanced adaptability of robots to humanity and environment is highly required in the aging society and in the symbiotic future with natural environment (see sidebar).

Waseda University established the Humanoid Robotics Institute in April 2000 to promote research that aims to construct a new relationship between humans and machines in an advanced information society. As researchers, we expect that in sometime in this century robots will provide housework assistance for the elderly, a well as entertainment and other functions to improve the quality of life for humans. To make possible this symbiosis between humankind, robot, and the environment, we need to build accommodations into a house's structure and functions

MACHINE LIFE

MACHINE EVOLUTION
Perhaps on other planets, organisms have evolved into machine life, cyborgs, or an entirely new form of synthetic life we cannot as yet comprehend. Indeed people have seen aliens that have appeared to be machine like with body armor or complex space suits.
These two publications are presented jointly here because they create a bridge between two generations of artists who nonetheless share concerns such as abandoning a strictly instrumental use of technology to embrace random phenomena and simulating human perception. Representing this first generation since the sixties, Norman White, a professor at the Ontario College of Art (OCAD, formerly OCA) in Toronto, Canada, taught in the seventies and eighties with his colleague Doug Back to many artists from the subsequent generation (the eighties), including David Rokeby. In addition to the author contributions, the catalogues for these exhibitions include multimedia complements (a CD-ROM for "Machine Media" and "Norm’s Robots" as well as a DVD for "David Rokeby"), along with video excerpts to complete the documentation of each of the works exhibited.

The catalogue for "Machine Life" brings together three author contributions. In "Norman White, Beginning," Ihor Holubizky looks at the remarkable critical interest sparked by media arts between 1968 and 1970, a period when White’s work was first emerging. White took part in some of the major exhibits organized in the United States in the late sixties, including "Some More Beginnings: An Exhibition of Submitted Works Involving Technical Materials and Processes" in 1969 at the Brooklyn Museum (New York, U.S.). Holubizky follows White’s career from the artist’s first robotic works of the seventies to his recent projects that accentuate the playfulness of his work. The author presents White’s sometimes ambivalent views on the difficult relationship between art and technology. He concludes by underlining the entropy found in White’s work, which distinguishes it from many artistic projects modelled on the notion of technical or scientific progress. In "Taken with Surprise," Caroline Langill points out that like White, artists from the subsequent generation were interested in technology’s unpredictability. When computer units of a media artwork exchange data randomly, the end results produce a range of varied experiences for viewers. In "Encountering Machine Life," Jan Allen, like Caroline Langill, stresses the role of entropy in White’s works, which, according to this artist, "represent unusable experimental models." The artist’s process may tend toward a form of productive failure in which technology frees itself from the uses predetermined by its functions. Allen explores the years of creative exchanges between White and his former colleague Doug Back from the Ontario College of Art. The two artists shared an interest in manual work that resulted in the construction of technological components for artworks as well as a decompartmentalized approach to interactivity. A description of works presented within the exhibition follows.

The catalogue for "David Rokeby" includes two author contributions. In "Between Chaos and Order: The Garden and the Computer in the Work of David Rokeby," Su Ditta relates Rokeby’s work to a garden, a space inciting both action and contemplation. She tracks Rokeby’s career path, describing emblematic works created since the eighties. She separates projects exploiting sound and language (Very Nervous System [1986-2004], The Giver of Names [1991-] and n-cha(n)t [2001]) from projects concerned with the boundaries between computer vision, controlled by a function of analytical observation, and human vision (Watch [1995], Seen [2002] and Taken [2002]). Other works (Machine for Taking Time [2001] and Steamingmedia.org [2002]) are linked to complex devices in which a real site coexists simultaneously with several representations of the site captured at different times of the day or year. In "Interpolation: The Method of David Rokeby," Sara Diamond reports on Rokeby’s technological and aesthetic research while evoking artistic projects situated between technical application (software, protocol) and artistic intervention. Diamond emphasizes the concept of invention in Rokeby’s work and the manner in which he himself creates the technological tools required to produce his works so that these tools then exist independently (Very Nervous System, The Giver of Names). The author analyzes this last work from the perspective of a theoretical reflection on the differences between strictly human faculties and computer functions for capturing and analyzing data. Finally, Diamond comments on the artist’s use of sophisticated surveillance tools whereby the immersion in the image and the aesthetic experience are often accompanied by a paradoxical update of the technology’s essentially coercive functions

MECHATRONICS FUTURE

I think mechatronics generally begins with mechanical design. That’s just my perspective, it may differ in your experience. If its a power window in a car, a hard disk drive platter machine, a blender, an amusement park ride or display, a surgical robot, whatever. They all start with mechanical design, performance goals and boundary conditions that are required for the mechanical system to be useful. This is why there needs to be great emphasis on the design of software tools that are extensions to the 2D and 3D CAD products that are currently available. Obviously, if you are engaged in mechanical design, you are in a unique position to the final outcome of the design project. The mechanical design work sets the boundary conditions of what is possible. If the design is all steel and heavy components, then the speed and throughput of the design will be limited. But the same design work in aluminum will be 1/3 the weight and it will be possible to increase speed and acceleration at the same time as cost for the motor drive will be reduced. Seems counterintuitive, but it works. But the it doesn’t end with mechanical design and that’s where we all get tied up. Is there a definable process for doing this kind of design work? There are some commonalities in the machinery building community, but there are wide variations at the same time. So for software companies, it gets harder to go to the next level to create value for their customers. The design process is iterative. So there’s a clue. What can we do to speed up the design process and improve the outcomes? Simulation. Take the output of the designed components and apply animation rules to the 3D solids which are very well characterized in the software. By using computer technology to dynamically simulate the behavior of the design before its built. Just like the prototyping process, you can find out a lot of really useful information while the simulation is running, early in the design cycle and without the cost of building and exercising prototypes. Mind-blowing! You can do “What-if” all day at low cost and in very little time. So this is clearly the way to go. And engineering software companies are currently engaged in process of creating these products. But the fun is only just beginning. What happens when the products we create have to be manufactured? There are a number of complex issues in the manufacturing and product documentation realm that become very complex without software tools to help with the tasking. New software products are being created to help integrate wire harness integration in automotive assemblies. Circuit boards can be modeled ast 3D solid objects for the purpose of integration with packaging.