During our initial robotics lab activity on Monday, January 20, Angela Bray, Phil Sommer, and myself, built a computerized Lego car fondly nicknamed “Rover.” As we familiarized ourselves with Rover and the LabView program, we experimented with different functions and their compiled reactions with our robotic car. After successfully programming Rover to drive in a straight line while playing a cheerful tune, we moved on to testing the wheel power functions directed to Port A and Port C on the Lego car. It was at this time that we discovered that varying the power to each port caused the car to stray from its head-on direction; by programming one port to output a higher power level than the other, Rover’s wheels turned in a circular direction.
Much to our pleasure, this week’s lab assignment involved experimenting in this manner with the port power levels. In order to successfully complete one of the tasks, the power levels had to vary enough to produce a diameter exceeding 2 feet within the circle traveled by the car. Using our knowledge of power variation from the previous casual trial, we succeeded in creating a large circle by programming one port to output a power level of 50 and the second to output 25.
The second portion of our robotics activity required us to measure the distance traveled by Rover given our inputted power levels, and to then compare the results with the calculations given by the LabView program. In order to properly determine the mean distance between the ruler and the program’s computations, we were provided with the fractional error equation:
fractional error = | druler - dprogram / [(druler + dprogram) /2] |
We completed 4 different trial scenarios for this experiment, starting with the lowest speed and increasing the levels, and each only running for 1 second. For the first trial, we programmed the power levels for both ports to output a speed of 25. Upon compiling the program, Rover jolted forward and stopped at 6 centimeters according to our measurement. LabView calculated that Rover had traveled 5.617 centimeters, and that his wheels had turned a third of a full rotation. The calculated fractional error of this trial was 6.6 centimeters.
During the second trial, we increased the outputted power to 55 and watched as Rover sped along and stopped abruptly at the 16-centimeter mark, which matched the distance results provided by LabView. This time, Rover’s wheels completed 95% of one full rotation to reach this distance.
By the third trial, we began to notice a developing trend in the relationship between our distance measurements and the LabView calculations. When Rover’s power level was increased to 75 (with the wheels turning 1.47 times), his higher speed caused his driving directions to become skewed as he veered to the left, causing us to grow concerned with the accuracy of our measurements. According to our results, Rover’s wheels stopped at the 23.5-centimeter mark, while LabView demands that the car stopped at precisely 25 centimeters. The fractional error between the ruler and the program resulted in 6.18 centimeters.
The following and final trial, with a new power level of 80 to both Port A and Port C, proved to be the most difficult to successfully accomplish as Rover drove in a curve away from his designated track; annoyed by this experiment complication, we repeated the trial and measured that the car had traveled approximately 28 centimeters using 1.55 wheel rotations. However, the results did not comply with those provided by LabView, which stated that Rover had only driven 26 centimeters. This trial yielded the largest fractional error of the experiment at 7.4 centimeters.
As we increased the outputted speed, our trials differed further from the calculated results of LabView and our measurement accuracy became compromised by the complications in efficiency. In conclusion, we determined that a higher power level produces a greater distance, yet with the higher speed comes the skewed direction of the car and the risk of collecting incorrect data.
I like that your blog is very detailed and takes you step by step through the process of building the robot and doing the various tests.
Rover’s looking good in that picture! Is there a way you could send it to me so I could use it? pssommer@suffolk.edu