Episode 8 Contest: How good is a Roomba?

[rcross]

83.4% ryanecross [at] gmail dot com

[lancelot]

The whole area can be divided in a square made of 10x10 smaller squares (most of them are visible, the others blurred by the floor motif)
During the 20 s the roomba works in the video, it covers ~8 of these small squares, therefore 8% of the surface. If a linear progression is assumed, the robot will cover 60% of the surface during 150 s (2.5 minutes).

Of course this assumes a homogenous distribution of "dirt" and that the floor motif has no influence on the roomba's behaviour - both assumptions are quite unlikely, therefore there is room for refinement in my own "probabilistic model" ...

[ringo42]

64.2%

[montesns]

I would say 93.3%...
Carlos Montesinos (robotics.phd@gmail.com)

[KerJiun]

86%

[epokh]

I can't see the video also in youtoube it says "no longer available" weird.

[RobotsAustralia]

Hmm, I guess 44% in the 2 1/2 minutes, calculation is real back of the envelope stuff.
cheers,
Murray
robotsaustralia -at- gmail dot com

[balle]

I would guess about 62%

[R2-D2]

61,314159265358 %

I will try to share with you my uncertainity calculations
a) on the video we can see that the roomba has the "extra much dirt" feature on, which makes him circle and therefore loose much time.
b) from the previously posted video on the forum about the roomba contest, we can see that the roomba achieves 63% of cleaniness.
however

this is a single shot measure, no stats.
there were obstacles
the arena size was larger
the given time was 7 minutes not 2.5

c) there will be an error of measure by the clever, but not perfect podcast guy's, let say +-1%
d) all the previous posts in ths thread

I suggest to repeat the cleaning measure as many times as there are posts. most of us will win once then. As there are not that many books to win, an average will be ok

[HKou]

42 should be the right answer but I expect the robot to be more effective, therefore 100 - 42 = 58% is my optimistic guess.