Top Challenges in Mobile Autonomous Robotics

[vidam]

I'm putting together a brief at work on the current challenges to Autonomous Mobile Robotics and I would appreciate any feedback on what you think they may be. And perhaps also you may refer me to the names of researchers addressing the challenges. I would like to get more information on their research or where the current focus is in robotics research.


Just a short, brief list is fine. I can do most of the leg-work myself.

Thanks much,

[ioannis]

Energetic autonomy is a newist area of mobile robotic autonomy, see the work of Stuart Wilkinson and Ioannis Ieropoulos, also the classic navigationaly autonomy problem is still in the works and probably one of the most important areas in mobile robotics.

[AndrewM]

Not sure what you will find in the way of researchers addressing the issue, but the primary problem I see with mobile autonomous robots is the lack of usable sensory input.  Example, a camera is great, and we can even find a given color or shape with that camera, but not accurately enough, and little beyond that currently.

I would take a look at the problems and solutions presented by entrants in th DARPA challenges.  Lots of good information on autonomous vehicles, their challenges and methods being used.

[ioannis]

It is true that imperfect sensors and actuators make autonomy difficult to achieve it is one of the main problems in navigational autonomy along with dynamically changing environments which the machines cannot adapt to fast enough. One of the ways they are dealing with the problem is evolutionary computing, genetic algorithms that allow the program to change and adapt accordingly, its biologically inspired and it seems that most creatures have developed perfectly with that techqinue, look into that too.

[maverick monk]

I would say that the biggest difficulty is reliability. to be autonomous the robot must be durable enough to not need repair. one of the biggest problems with the darpa bots is so many of them broke befor completing the course. the more complex the control system, the more stuff there is that could break. just think, a 10 second lagg on a computer can be annoying, on an autonomous bus a 10 second lag could cost many lives

[Admin]

Just pick up a robotics journal and read the table of contents.

and a list here too:
http://www.societyofrobots.com/robotforum/index.php?topic=604.0

[vidam]

Thanks for sharing the link RI Seminar Vodcasts for us.
http://www.cs.cmu.edu/~robocast/