Human Emulation - Data Analysis
FORWARD WALKING TEST - ANALYSIS
These charts graphically display the results of the straight
line walking test for robots we tested. If the robots
wandered to their right, the amount is shown as a positive
amount, and wandering to their left is a negative amount.
The fact that there is not a clear pattern indicates that our
test platform is close to level, and our home position on the
robots are reasonable. In a shocking coincidence, for Test
15 on the Robonova-1 and Test 4 on the I-Sobot the
robots landed exactly on the center line. If you watch the
videos for these tests, you will see that it is a random
event. Remember, we did not make any course
corrections during the tests.
NOTE: These robots are not meant for children. All of the humanoid robot models discussed on this website use
powerful servos that can pinch or even crush your fingers if you are not careful. Consult the recommended minimum
age from the manufacturer before allowing children to use any of these robots.
It was not a surprise to see there is a strong correlation between the number of steps the robot took, and the time necessary to complete the
test, as shown below. The bigger surprise is that the variance from center is not strongly correlated to the time necessary to complete the test.
What we learned from this is that while the I-Sobot and the Robonova-1 each took roughly the same number of steps to complete the test,
the I-Sobot was twice as fast. This means that each robot has roughly the same length of each step (relative to its body height), but the
I-Sobot was able to perform so much better because it has a continuous walking motion. On the other hand, the KHR-1 took substantially
fewer steps, which indicates that each step is proportionately longer (relative to its body height) than the other robots. This gave the KHR-1
the edge needed to record the fastest time.
180 DEGREE TURN - ANALYSIS
The charts below show the variance side-to-side against the variance front-to-back for our Human subject, and all the robots we tested.
The chart on the left shows the raw results measured in inches. Since the robots are so much smaller than the Human, you would expect
the variance in inches to be smaller. Therefore we have also included the variance as measured by the % of Body Height in the chart on
the right. So when the Human attempted to turn 180 degrees, it took about 23% of its Body Height, but the Robonova-1 required a
whopping 72.8% of its Body Height to make the same move, or roughly 3 times worse. The I-Sobot gave similar results requiring 73.3% of
its Body Height to turn 180 Degrees. The KHR-1 performed the best, requiring only 41% of its Body Height to turn 180 degrees. The chart
on the right graphically displays the performance difference based on the lack of rotation at the hip or waist for the robots we tested. We
plan to test other robots in the future which have hip rotation so that we can measure how much better the performance really is.
|FREE Robot Test Drive
Would you like to try out the
I-Sobot or the Robonova-1 for
yourself? Click here to sign up for
a FREE Test Drive from the
comfort of your own home.