Researchers from across Europe are being trained in Genoa, Italy, this month in advance of taking possession of their very own iCub: a robot designed to have the physical and sensory capabilities of a two-and-a-half year old child.

For the researchers involved, one crucial characteristic of the new robot is that both the hardware and software are open-source and designed for easy collaboration. Whether the researchers build better cognitive architectures, learning algorithms, sensors or limbs, once their work has been proved on the European Commission-funded iCub, it can be shared and used to improve the next generation of machines.

The finished iCub stands about as tall as a three-year-old and is designed to encourage interaction

Though not the first open-source robot, iCub underscores a trend that is poised to increase the productivity of artificial-intelligence (AI) researchers, in the same way the open-source movement has enhanced work in other sectors of design.

"Open-source is an open faucet in a desert," said Olivier Sigaud, a professor at the Institute for Intelligent Systems and Robotics in Paris.

There are two models for driving long-term progress in robotics: industrial and academic, said Giorgio Metta, an assistant professor at the University of Genoa and the Italian Institute of Technology who is among the leaders of the RobotCub Consortium. Industry is very well-organized and can devote effort and money to the field. Academia, though less structured, represents a large proportion of robotics researchers.

It is for the latter group that the open-source approach is so important.

"I'm perhaps too enthusiastic about it, but I see only advantages: better collaboration, reproducibility of results, shared debugging, faster improvement, etc.," Metta said. "For theoretical research, standard publication in scientific journals and at conferences is OK. For robotics, sharing code is better."

After Sony retired the Aibo, enthusiasts worked to build a new robot with open-source underpinnings. The New4LR (four-legged robot) is ready for RoboCup soccer

The conventional approach of using commercial robots has long caused problems for academics. They must sign limiting nondisclosure agreements to gain access to proprietary technology, and they may be prevented from going into low-level control code. Then there's the headache of discovering that a line is being discontinued. When Sony abandoned robotics in 2006, for example, many groups that had been using the company's Aibo robot as their primary research platform were left in the lurch.

Opening up AI

Open-source robots are not new. Long before Lego Mindstorms debuted 10 years ago as a kit available to the public, its predecessor was regularly used for teaching and research in universities across the globe. The open-source culture that grew up around the technology--not to mention its sheer flexibility, ease of use and low cost--made it extremely popular among academics.

Another project, OpenPINO, was more ambitious. A team at Sony Computer Science Laboratories, led by Aibo developer Hiroaki Kitano, created an open-source humanoid robot intended for the academic community. Unfortunately, OpenPINO did not reach many researchers, particularly outside of Japan. Whether that was for technical or price reasons is not clear, but Sony's subsequent departure from the field makes it a moot point.

Rather than focus on building robots, some researchers have tried to build systems that allow software, once written, to be shared across machines. One example is Tekkotsu, created at Carnegie Mellon University in Pittsburgh.

With 53 degrees of freedom, the iCub lets researchers test their theories about interaction and learning in humans

"Tekkotsu is intended to make it easy to develop sophisticated applications on mobile robots, by providing an extensive set of well-integrated primitives for perception, navigation, manipulation and control," said project head David Touretzky, a research professor of computer science at CMU. "We are seeking to change the way undergraduate computer science students are introduced to robotics. We believe that the best approach is to give students technically sophisticated robots and good high-level software tools, and then teach them how to use these tools creatively."

The Tekkotsu project earned Technical Innovation Awards in 2006 and 2007 at the Mobile Robot Competition and Exhibition, held during the annual meeting of the Association for the Advancement of Artificial Intelligence. Ironically, much of the initial work was done on the Aibo.

For instance, Spelman College used Tekkotsu on the Sony robot to compete in the RoboCup competition, and some of Touretzky's own students got an Aibo to act as a blackjack dealer. The Tekkotsu researchers are now working to support more hardware platforms and even to build their own hardware.

Another new open-source robot, the New4LR (four-legged robot), has been designed precisely to fill the void left by Aibo's departure. Priced at just under $5,500, the robot should be affordable for most universities. Its open-source underpinnings are among the ways the robot will contribute to the field, said Oskar von Stryk, a professor of computer science at the Technical University of Darmstadt, Germany, where the New4LR was developed.

"Modularity and reusability are required to enable the technological evolution of autonomous robots," von Stryk said. "In the long run, this is only possible with open software and hardware modules that enable an unlimited number of researchers and developers to share their particular contributions."

Enter the iCub

Beyond taking an open-systems approach, the iCub robot is designed around a very particular take on artificial intelligence. The designers eschewed a top-down approach to engineering specific desired behaviors, a strategy often identified with Japanese robotics. Neither did they design knowledge-based algorithms that are entirely abstract and are supposed to work regardless of the robot used (and that fail to address basic control and perception issues).

Instead, the iCub's design has been based on the so-called embodied approach.

The robot learns skills through exploration and interaction with its environment and with other agents, particularly humans. It learns about the world through its experience of what its body can do and how its sensors help it accomplish specific tasks or goals.

To learn in this way, the iCub's cognitive architecture is based on what psychology and neurology have uncovered about the innate abilities of human newborns and their subsequent development. A "modulation circuit" enhances and refines existing skills by combining a memory of past sensor input and actions with the motivation to achieve or optimize a particular task.

A separate circuit lets the robot "rehearse" possible sets of actions and, based on its associative memory, see the expected results. Based on how successful those plans turn out to be in simulation, they can be kept, thrown out or amended. The expected outcomes can be compared with those that occur when actions are actually performed.

For the robot to learn like a human, it must have the functionality of a human. To that end, its designers say the iCub will be able to crawl on all fours and to sit up. Its hands are said to be able to perform dexterous manipulation, and its head and eyes are fully articulated. It has visual, vestibular, auditory and haptic sensory capabilities. In short, the iCub's strength, range of motion and senses are as close to those of a toddler as possible given today's technology.

The middleware for the iCub is a software architecture called Yarp ("yet another robot platform"), which supports modularity. New devices, functions and communications channels can be used by defining an interface that controls the device at a low level but allows standardized high-level commands to bring the devices into play.

According to the iCub team, this will encourage the long-term reuse of successful modules and the replacement of those that are less successful or that have been superseded, producing a kind of robotic "gene" that could live for many generations of bots.

The learning curve

However sophisticated, iCub is not intended to be a finished product but is designed as a platform for further research. To that end, the RobotCub Consortium has sent out calls for research proposals, with robots going to the teams that posit the best proposals. In the first wave, the top three proposals came from Imperial College London, Pierre and Marie Curie University (Paris) and Lumière University (Lyon, France).

The Imperial proposal was submitted by Murray Shanahan of the computer science department, upon whose work some of the iCub cognitive architecture is based, and Yiannis Demiris of the department of electrical and electronic engineering. According to Demiris, the team intends to design and implement a cognitive architecture that will let the robot explore and learn to manipulate objects in its environment. The scientists favor models of the "self" and the environment that enable mental rehearsal of possible actions.

In particular, Demiris will be concerned with the learning. "I would like the iCub to have the capacity to [discover] the effects of its actions on objects," he said. "This learning by experimentation treats the robot as a little scientist, capable of forming hypotheses about the world and experimenting to confirm or disprove them." He will also endow the iCub with the ability to learn by imitation.

In Paris, Sigaud at the Institute for Intelligent Systems and Robotics said his project is based on a simple scenario. "The iCub robot will sit at a table with a few objects within reach, and interact with these objects according to its own drives," he said. "Then a human user will modify the robot's behavior so that it meets simple constraints. The user can do this both by demonstrating the expected behaviors and cheering or rewarding the robot, depending on its actions."

Peter Ford Dominey of the Laboratory for the Study of Machine Cognition in Lyon is also concerned with robot-human interaction, in the form of cooperation. "We have demonstrated that robots can learn to perform novel, non-prespecified, shared, cooperative tasks, alternating their actions with the human's, but the robot does not 'understand' the real goal of the activity," he said. "In comparison, we know that humans possess a profound and likely innate propensity to understand and communicate goal-related intentions." By 18 months, he said, "children understand the intentional goals of adults demonstrating cooperative games."

Through the use of what Dominey called "situated simulations," the Lyon team's robot will "acquire knowledge about the results of actions that will begin to provide a more profound sense of 'meaning,' including the ability to understand the intentions of others," he said.

All of the researchers agreed that the open-source path is important. "Commercial software requires and imposes a wall. One has to stop, worry about how and how much to pay, [then determine], 'Does it run on a second computer? Can I have a student use this?' etc.," Sigaud said. By contrast, "The iCub open-source simulator can be downloaded quickly and freely, and anyone from a clever high school student to a seasoned researcher can begin to use it immediately, to share their results and to benefit from the collective effort of the surrounding community."

Three other successful proposals to the consortium came from the Technical University of Munich (Germany), Pompeu Fabra University (Barcelona, Spain) and the Middle East Technical University (Ankara, Turkey).

—Sunny Bains is a London-based scientist and freelance technology journalist.