Important legal notice
Contact   |   Search   
Headlines Published on 27 September 2004

Title Smarter, cognisant robots not far away now

As outlandish as it sounds, in the not-too-distant future, robots will be able to perform, with an increasing degree of autonomy, some of the more tedious and dangerous tasks we delicate humans care little for. Young scientists in the EU are paving the way for this I, Robot sounding scenario.

Could I, Robots safely do the things humans cannot © Image: PhotoDisc
Could I, Robots safely do the things humans cannot
© Image: PhotoDisc
In the near future, automated robots will help us in a number of different fields, such as transport, cleaning, mining and agriculture, according to a young Swedish scientist whose dissertation on behaviour-based mobile robotics is causing quite a storm. These, dare we say, ‘intelligent’ robots will steer themselves without human involvement and will even be able to plan their own successive tasks. The potential applications for this burgeoning technology are endless, and it can help save lives and protect the environment in the process.

To learn from their mistakes – the key to a thinking and forward planning machine – next-generation robots must be able to detect if something is wrong and even find out what caused it. Today, the most common way of doing this is with models that predict the robot’s position after a certain action. If the robot ends up in the predicted position, it is acting correctly. But if it is off course, then something is wrong, and this is where the difficulty lies, asserts Ola Pettersson of Sweden’s Örebro University, author of the paper.

Indeed, experts concede that setting up precise models to map this kind of activity is virtually impossible. So Pettersson decided to study how mistakes can be discovered using behavioural patterns. “We will simply be able to tell the robot that it’s doing something right or wrong, instead of describing it in mathematical models,” says Pettersson. But he cautions that studying behavioural patterns to detect mistakes in mobile robots is a new approach, which still needs to be properly tested. If successful, however, the method will make it easier for humans to use robots.

Trained to serve
The paper, entitled ‘Model-free execution monitoring in behaviour-based mobile robotics’, shows that it is possible to detect a mistake, and even to determine what type of mistake it is, by studying the robot’s patterns of behaviour. Several different techniques have been tested, and their usefulness has been shown in numerous experiments, using both simulated and real robots. Statistical methods have been used to compare the various results.

Pettersson is one of a group of young scientist in Europe who are pushing the frontiers in this exciting field. This is an encouraging sign for the EU, which is fully behind promoting the careers of young scientists, engineers and technologists. It has set aside significant sums in successive research Framework Programmes for human resources development in the sciences. It has also set up a dedicated budget in the ‘Future and emerging technologies’ (FET) initiative within the Union’s ‘Information society technologies’ (IST) programme to explore the scientific realms ‘beyond robotics’.

In the current Sixth Framework Programme (FP6) for research, several projects have already got underway this year, including the setting up of a European Robotics Research Network (EURON) and the FET project called ‘The Cognitive Robot Companion’ which is investigating the perceptual, reasoning and learning capabilities of robots. An interesting list of research endeavours by other young European scientists is presented on the EURON website.

Source:  Swedish Research Council, IST, EURON

Research Contacts page

More information:

  • Örebro University (Sweden)
  • FET Proactive Initiative on Beyond Robotics (IST programme)
  • Robotics in Europe  
  • EURON (homepage)
  • Other interesting theses in robotics (EURON)
  • Human resources and mobility (Europa website)
  • I, Robot (on IMDb)

  •   >> TODAY'S NEWS