HIS Western colleagues joke that Hiroaki Kitano must be one of identical triplets, because no single individual could do as much as he does. He is simultaneously a researcher at Sony鈥檚 Computer Science Laboratory in Tokyo, an adviser to Sony president Nobuyuki Idei, and the leader of the Kitano Symbiotic Systems Project, funded by the Japanese government with the modest aim of reverse-engineering life.
Kitano is also the instigator and main driving force behind RoboCup, the world robot soccer championships, which this year is being held at the Melbourne Exhibition Centre from Monday to next Sunday.
Through founding RoboCup, Kitano hoped to speed the progress of robotics research. Three years ago, when the championships were first held, merely finding the ball was a big ask for a robot. Since then, 鈥渢he technological progress has been pretty staggering鈥�, says Andrew Jennings, chair of RoboCup 2000 and a professor of computer science at the host organisation, RMIT University.
Advertisement
Thirty-five teams from 12 countries participated at those first championships in 1997 in Nagoya. This year, more than 100 teams from 35 countries are taking part.
Australia is represented by strong teams in every division. In the small robot league 鈥� toy-sized machines no more than 18 centimetres high 鈥� the University of Queenland鈥檚 RoboRoos came second in Stockholm in 1998. Their placing was matched in 1999 by the University of New South Wales team, UNSW United, in the legged (as opposed to wheels) league. And in the simulator league 鈥� played entirely on computer 鈥� the CISRO Cyberoos have done well in recent European competition.
Over time RoboCup has developed in unexpected ways. One surprise was the enormous popularity of the tournament with the general public as well as participants. Another was the extent of the educational experience.
The ultimate goal of RoboCup is to take on the Soccer World Cup champions of 2050, and win. To this end, a humanoid robot competition will debut in 2002.
But the idea of humanoids beating humans, if only at sport, prompts visions of a future in which self-replicating terminators overpower people. Such concerns have been around for a long time in the West. In Japan, by contrast, robots have always been seen as helpful assistants for humans, not as replacements.
Kitano himself argues that no-one has yet figured out how robots might self-replicate. He also points out that a machine whose intellectual performance rivals that of people would have 鈥渁lmost no market value鈥�.
On the other hand, there is a definite need for robots designed for specific tasks, especially those which are dangerous for humans. Last year, Kitano dreamed up the idea of applying some of the technologies developed for RoboCup to disaster rescue.
His stimulus was the 1995 Kobe earthquake, which killed more than 5000 people. Many of them could have been saved if rescue efforts had been quicker. Jennings came up with a local variant on the theme 鈥� robot lifesavers, or 鈥渓ifebots鈥�.
Though the surface of the sea is not as difficult an environment as a city after an earthquake, finding someone in the water can still be a challenging task. So, the lifebot 鈥� a modified life raft which can be grabbed by a drowning person 鈥� is equipped with an array of sensors, including sonar, and a camera to allow remote operation.
Much to Jennings鈥� surprise, there are commercial applications for lifebots. Customs officials are interested in using them to film smugglers dumping contraband from their boats. And local authorities are looking for cheaper ways to patrol beaches. Also, according to Jennings, human lifesavers are receptive to a helping hand from technology. How very Japanese of them.
