Brown Bag Lab Lunch Series

DESIGN COMPUTING RESEARCH FORUM

return to schedule

add a new review

Reviews of Distributed Multi-Robot Exploration and Mapping:


  • John Christopher Mayfield

    I remember back when Eun Soo gave his presentation involving AR, I thought "Well, it has a camera, a computer and a movement and orientation sensor, right? So couldn't we create computer models of buildings by just walking through them?"

    That was a stupid thing to think. =)

    I'm consistiently amazed by the number of problems that I had no idea even existed which have already been solved in truly ingenious ways. This talk was a great continuation of that trend. It is staggering how many issues are involved with orienting the robots within a space and insuring that tiny errors in turning and distance sensors do not add up to create warped maps. What is even more staggering, though is that many of these issues have already been dealt with.

    When I first saw Aibo, I thought "What retard thought that would sell? You can get real dogs for free outside the grocery store." That was a stupid thing to think. Obviously the value of Aibo is not in it's ability to emulate the animal it resembles, but in that it can be programed to respond to just about anything in any way the programmer wants--even to the point of getting Aibo to play soccer. After the explanation of all the problems involved, I considered that one blocked shot to be the most impressive thing I'd seen in quite a while.

    edit this review


  • Scott Brandon McDonald

    Robots playing soccer?.. its an interesting competition to test the skill of programmers. The AIBOs are neat little robots and I expect we will see more of them in the years to come. The Multi-Robot mapping problem seems a bit more difficult, but I was impressed with the abilities the current robots have at mapping and exploring. I hope that someday the robots will do all the tedious surveying work that is necessary before starting a project.

    edit this review


  • John Hilgeman

    ahhhh, AIBO. I was introduced to the little pooch in this class last year and he/she seems to keep reappearing on the technology research scene. At first, I thought AIBO was one of these silly plastic walking toys that have been around for years. However, when I saw it move and react I was very entertained. I was still skeptical of this particular doggy becoming a sensation - I thought it might spark a new trend in robotic research - but it seems AIBO is the new trend, what with the soccer tournaments, and the waking up the master, and the reading email, and what not...

    edit this review


  • Eithon Michael Cadag

    Very cool presentation by Prof. Fox. I particularly liked learning about the mapping abilities of the robots. I've always wondered exactly how robots can create maps of their environment, especially given friction that could alter their vision of the environment. I thought the way that the robots closed loops to correct for their vision of the environment was very interesting and educational.

    While the robotics aspect may not have any direct application to design computing, its possible that the algorithms and AI used in robotics could be used in CAD software, perhaps in a meshing program to combine building components from different drafts into a single one?

    edit this review


  • Rosanne Weiling Chien

    This presentation was very fun, and I think even attending the different contests would be very rewarding because you can actually see your work in action, and see how well (or perhaps not so well) it actually works when compared to other models. Yet seeing these robots perform human measurements to a great deal of accuracy (to the point where we think the robots measurements are better than the judges measurements) is truly amazing. Even for us to have that much accuracy would require other tools (rulers, protractors, etc) and could deem difficult and arduous. Having a robot perform these measurements relatively fast is great.
    Yet, the idea of having robots do the exploration and mapping that is commonly, and at times, currently, done by humans proves very beneficial. There are many places where human exploration is not possible or just not safe, and having a robot do this instead is great. Another idea would be to decrease the size of the robot. Generally, it is difficult to decrease the size of any technology, but having a smaller robot can also prove beneficial. There are many areas that are difficult for humans to explore because we are just too big. And if we are an even super small robot, perhaps we can do "medical exploration" on ourselves.

    edit this review


  • Julie Dawn Pravitz

    This presentation was great! I found both the Robocup and the robotic mapping amusing and facsinating. Regarding the robotic games, I had no idea of how complicated the programming was for the "players". But it all makes perfect sense. Something as simple as figuring out one's location on field or judging the distance of the ball is actually very complicated. The methods the robots used to figure out these things surprised me; relationships between objects, the size (number of pixels) of the ball, the speed in which the ball increased in size to judge the distance... wow! We humans take forgranted our ability to make simple visual judgements!

    The robotic mapping was also great because of it's usefulness. It's faster than human measurements and more accurate. As the technology becomes more inexpensive, the uses of these little robots will dramatically increase. They could make existing conditions surveys required by architects on modernization jobs much quicker. It would be great if they "drawings" produced by the robots could be converted into a cad file. That would be the ultimate tool!

    edit this review