June 28, 2008

Fatima Sajid gives us an insight into development and the technological claims by the producers of the first to the fourth generation of robots.

Humankind’s untiring efforts to explore and study worlds other than ours have given rise to technological breakthroughs in space travel and probe on the unfamiliar and far-off terrains. Recently, Nasa successfully landed its Phoenix Mars Lander on the arctic region of Mars. The main purpose of this latest mechanical visitor from earth to the red planet is to dig deep and explore the frozen layers or water ice underneath.

But such explorations have given rise to man versus robot debates, like are robots efficient enough to perform a task in a smarter way than humans? Are robots better explorers than humans. Perhaps the answer lies somewhere in the middle.

Robots have proved better than humans in situations which are too tedious, too dangerous or too difficult for a person to perform in. They can repeatedly do the same chore assigned to them without boredom or fatigue, above all a good point as compared to humans is that they do not need coffee breaks and lunch meetings.

After all, robot has gone where no man has gone before since it does not need the regular essentials like breathable air, water and proper shelter to survive. It can survive the harshest of temperatures and obscure terrains as space explorers. However here on earth, in almost every aspect, machines are indispensable for example in factories, construction sites and most importantly, in the field of medicine, they have proven their worth.

Space explorers

For decades now, robots have been pioneering space exploration as planet Earth’s first emissaries to other worlds. The twin robots on Mars have exceeded the expectations of their designers and engineers. And now, the latest mechanical visitor from earth, the Phoenix Lander, has already begun unfolding mysteries regarding the alien terrain of the red planet. It has a robotic arm, which can scoop up the ice and dust, a wet chemistry lab including small oven-like instruments, which will analyse soil to detect its compounds.

The success of the Lander has greatly encouraged space scientists, and according to Ed Sedivy, Phoenix programme manager at Lockheed Martin Space Systems Company, the entry, descent and landing phase of Pheonix, and its flawless performance has the scientists asking for more.

A robotic explorer’s duties are to dig into the ice, examine the compounds, visualise the terrain and also to send the data back to earth from a distance of 171 million miles, (275 million kilometres). But the question is, could a robot recognise and identify alien life?

In order to make the robot intelligent enough to differentiate between a stone and microbial life, a team of scientists from the University of Leicester used robots to study rocks here on earth. They collected rock samples, which were similar or closer to the kind of material that they would find on Mars. The lab results were promising because the robots were able to detect living organism in the rocks.

The next much-awaited mechanical explorer to visit Mars and literally dig for life is the European Space Agency’s ExoMars mission. Scheduled for the year 2013, the probe will consist - among other state-of the-art characteristics a rover - which will travel several kilometres on the alien surface. Another instrument will be the Urey Mars Organic and Oxidant Detector, funded by Nasa, will search for the molecular signature of proteins. Thus, once again, the search for ‘microbial Martians’, will be dependant on machines, and not humans.

Nasa’s Jet Propulsion Laboratory’s New Frontiers Programme comprises various stages and areas in exploring the cosmos through robotic probes. For instance, the Kuiper Belt, Pluto explorer will enable researchers to understand the controversial star and its moons including other rocky objects circling that specific region; the Jupiter Polar Orbiter with Probes will help in understanding the inner structure of this gas giant (which enabled the formation of the other terrestrial planets in our solar system), the Venus in Situ Explorer will give us a clearer picture of how climatic changes on the planet gave Venus its extreme weather patterns today and also lay the foundations for a sample mission in the future. And then, there is the Lunar South Pole-Aitken Basin sample return mission. It will obtain samples of the ancient deep-seated rocks, analysis of which will reveal information regarding the composition and history of the earth’s relation with the Moon and vice versa.
 

The robots have advanced mechanics to perform almost all tasks. It is predicted that in about four to five decades from today, they will be preferred over humans but the only thing they will lack, say scientists, is the ability to interact with the physical world and matters of life and death.
 

Generation next

According to Hans Moravec, Principal Research Scientist at the Carnegie Mellon University Robotics Institute, robots have four evolutionary stages so far.

The first generation machines have the intelligence to operate the task they are designed for which has the mobility and the ability to understand and manipulate what’s going on around it. It is referred to as the first generation universal robot. The tasks that these robots can undertake and have the intelligence to perform can be compared to what a small lizard can perform.

The second-generation robots will have a built-in computer 50 times more powerful than the first generation robots. Hans compares its performance to picking up an object, which might be part of some big job, could be done with one hand or with another hand of the robot. The robot will be able to analyse and recognise where it has gone wrong and correct its action accordingly and learn through it. For example, the robot has a module that is responsive to collisions while it does its tasks and gives a signal that ‘something bad has happened’. If the task assigned has been completed accordingly, it will give out a signal that says, ‘something good happened’.

The difference between the first and second-generation robots can be marked in reliability and flexibility in performing the tasks. “Take a first generation robot that’s putting away dishes, explains Moravec, “Maybe it’s programmed to always grab certain objects in a certain way. And has motion planning and collision avoidance. But still, some things may have been overlooked in the programme. Whenever it goes into a certain cabinet, it ends up always catching its elbow in the door”. The second-generation robot, on the other hand, will use an alternative so as not to make the same mistake over an over again. Instead, the robot will tune itself. This second generation robot will be comparable to a small mammal due to its ability. The year of this machine will be 2020.

But the interesting generation is still backstage. This will contain a ‘world simulator’ which will allow the third-generation robot to make many mistakes in its mind, running through scenarios in simulation rather than physically, Moravec predicts. This makes it better at whatever task it is assigned to do and avoid mistakes by ‘thinking’ it out first, thus it is capable of getting it right at the first time. When this robot enters a room, it makes a mental inventory of everything in the room so as not to make any mistakes when it is physically interacting with the objects or persons presence.

For example, when this machine sees an egg, it has to know how much pressure will be required to hold it or how much strength will be required to hold each piece of cutlery. “Third generation robots should be able to deduce something about the internal state of the human beings around it, if a person seems to be in a hurry or if this person seems to be tired. You can probably deduce that from a modest observation of body language,” explains Moravec. This is something already being worked at, at Massachusetts Institute of Technology (MIT). With additional programming in the robotic brain, interactive dialogue can also be ingrained.

For example, if you ask it, “Why did you not go into that room?” and the robot would answer, “Because Bob is in there and I know he’s upset and my moving around him will upset him further”. It will also produce sounds like something out of a sci-fi futuristic movie, but not all that sci-fi, according to scientists. This third-generation universal robot has often been compared to a monkey.

Now we come to the real mind-boggling artificial intelligence stuff, a bit scary for some. The fourth-generation robot; this machine will have enhanced features than the third generation robot’s ability to ‘model the world’. They will be able to do many tasks, probably better than humans since really deep reasoning involves long deductive chains and keeping track of a lot of details. Human memory is not that powerful, says Moravec. Think Robin William’s character in the ‘Bicentennial Man’ to mind, with a few variations.

This kind of machines can be realistically very useful in finding and disarming bombs. In fact, scientists predict their abilities will be better at this than humans. According to them, a robot’s built-in radars and frequencies can be tuned for a particular task like seeing through various materials like walls and packages (without opening them).

The robots have advanced mechanics to perform almost all tasks. It is predicted that in about four to five decades from today, they will be preferred over humans but the only thing they will lack, say scientists, is the ability to interact with the physical world and matters of life and death.

“The kind of social intuition we have is very powerful and probably to close to the full processing power of our brain – the equivalent of a hundred trillion calculations per second”, continues Moravec, “plus, a lot of very special knowledge, some of which is hard wired, some of which we learned growing up”. However, the scary part is that the fourth generation robots will understand us better than we understand ourselves and might have the ability to even manipulate us.

Ray Kurzweil, the first developer of the initial omni-font optical character recognition and the first person to develop print-to-speech reading machine for the blind, including the winner of the 1999 National Medal of Technology, predicts that soon tiny robots will be implanted in a person’s brain and other parts of the body to improve mental and physical health. “It’s really part of our civilisation. But that’s not going to be an alien invasion of intelligent machines to displace us,” says Kurzwell.

The leading American inventor also adds that, “I’ve made the case that we will both have the hardware and the software to achieve human level artificial intelligence with the broad suppleness of human intelligence including our emotional intelligence by 2029.” He further clarifies that, “We’re already a human machine civilisation; we use our technology to expand our physical and mental horizons and this will be a further extension of that”.

And then, there will be the half-human, half-machine individuals passing us on the street or probably having a cool drink at the next table. These people will have robotic body parts in place of their actual parts that they lost in accidents or diseases. Or maybe even just to enhance the functions of the body. Scientists also predict that there will be a future in which intelligent robots and humans will work side by side. But it is even distressing to think for a second the possibility of their taking over us.