The Challenge of Robotics
Friday, January 30, 2004
Robotics Challenges - RoboCup
Okay - so what are the Grand Challenges in Robotics? The Japanese RoboCup movement has defined two challeneges robot soccer, and robot rescue.
Rescue is described in an AI magazine article which is well worth a read:
AI Magazine: RoboCup Rescue A Grand Challenge for Multiagent and Intelligent Systems.(use of robotic technology in rescue missions)(Statistical Data Included)
Both football and rescue challenges exist in physical and simulated guises. In addition, the football (soccer) leagues cover a range of robot sizes and morhpologies, from small tabletop robots through dustbins on wheels to the humanoid robot league.
There is quite a large movement towards getting the (search and) resuce challenge more widely recognised - for example:
Dr Robin Murphy's proposal to the CRA
not least because rescue robots have already been fdeployed fro real, even if with only limited success:
Center for Robot Assisted Search and Rescue
Here's some video footage from the World Trade Center wreckage
DARPA have got into the idea of Grand Challenges in robotics too, with the DARPA Grand Challenge, in which unmanned autonomous vehicles must drive between Los Angeles and Las Vegas. The $1 million prize will be awarded to the team who submit the fastest vehicle to complete the challenge, if any do, as long as it completes the challenge in under 10 hrs (how upset would you be if the single traffic light meant you missed it by a minute..?)
Historical Grand Challenges
So is there a history behind grand challenges, or are they limited to recent years? The following talk reviews some historical challenges, such as Huygens challenge to calculate how much hemp was grown on Jupiter (rather than how many Mars bars it takes to cure a case of the munchies...?) or the well known (perhaps, best well known) challenge of the 17th century - the solution to the problem of identifying longitude at sea:
Lessons from Grand Challenges
A rather more formal review, this time in the context of US National Academy Engineering report:
Concerning Federally Sponsored Inducement Prizes in Engineering and Science
offers a taxonomy of prize contests, both current and historical. (If you can't be bothered to read the whole thing, there is an executive summary of the report.
US High Performance Computing Challenges Exchange
This US site complements the UK one of the previous posting:
Computing Research Association Grand Challenges
A conference report suggests five challenges worthy of consideration:
Grand Challenges conference report
Systems you can count on
A teacher for every learner
911.net (ubiquitous information systems)
Augmented cognition
Conquering complexity
There are individual pdf descriptions of each of these challenges available too.
Grand Challenges in High Performance Computing
Grand challenges have a long(?!) and well-regarded history in computing research.
For example, the following UK site seems to act as a UK home for such challenges:
Grand Challenges for Computing Research
The site describes the essential features of a Grand Challenge as follows:
A grand challenge should be defined to have international scope, so that contributions by a single nation to its achievement will raise our international profile. The ambition of a grand challenge can be far greater than what can be achieved by a single research team in the span of a single research grant. The grand challenge should be directed towards a revolutionary advance, rather than the evolutionary improvement of legacy products that is appropriate for industrial funding and support. Finally, the topic for a grand challenge should emerge from a consensus of the general scientific community, to serve as a focus for curiosity-driven research or engineering ambition, and to support activities in which they personally wish to engage, independent of funding policy or political considerations.
Indeed, the site shows how the challenges have reached such a level of maturity that there are even criteria to judge the worth of proposed challenges:
Criteria of Maturity for a Grand Challenge
The criteria page also offers examples of completed and still open challeneges in high performance computing (e.g. a championship chess program, prove that P is not equal to NP, and a new one - a wearable computer that can act as a guide dog (?computer, surely) for the blind) as well as other sciences (for example, getting a man on the moon, or sussing out a grand unification theory in physics).
Competing for the Future...
Here's the original abstract
Competing for the Future: The Role of Grand Challenges in Robotics
Tony Hirst
Department of Telematics, Open University,
Walton Hall, Milton Keynes, MK7 6AA, UK
Grand Challenges play an important role in focussing research attention across many areas. One of the key drivers of artificial intelligence research in the 1960s and 1970s was the race to construct a computer program capable of beating first a human, then a human grand master, then the human world chess champion.
In the field of robotics today, there are several candidate Grand Challenges. As robotics research takes inspiration from across a wide range of disciplines - engineering, computer science, animal ethology, anatomy, and psychology, to name but a few - these Challenges have the potential to allow researchers from many academic areas to work together on a common problem and bring to bear expertise and practical understanding from their own discipline to a shared forum.
Based in Japan, the RoboCup organisation, which started in 1993, oversees grand challenges in robot football and robot search and rescue. International teams from academia and industry compete annually in a variety of leagues that define the size of the competing robots. Simulation leagues also provide an opportunity to try out high level intelligent software control systems unencumbered by the vagaries of having to control a real robot. The ultimate goal of the RoboCup soccer grand challenge is to construct a team of humanoid robot football players capable of beating the human world cup champions, by 2050.
The US based DARPA 2004 Grand Challenge requires an unmanned, autonomous ground based vehicle (such as a land-rover) to drive along a predefined route between Los Angeles and Las Vegas in under 10 hours. Unlike the 'amateur' RoboCup Challenges, the team that completes the DARPA Grand Challenge in the fastest time (if any) will receive a $1 million cash prize.
In this talk, I shall review the history of robotics grand challenges and identify the key elements that are required of a Grand Challenge if it is to provide an acceptable focus for research effort by the research community, as well capturing the public interest. I will also show how the softer biological and social sciences are just as important in the robotics endeavour as practical engineering and theoretical computer science.
