Performance Test 3 spans over 3 lab days. In these labs, we worked on various tasks involving the AEV, including running tests, recording data, and finalizing the Arduino code. Specifically, we worked on testing the third prototype, which we named Hermes. Testing this prototype leads us to determine what our final AEV model will look like, as well as finding a code that will be the most efficient.
A short video of one of our tests can be viewed here.
Performance Test 3
Continuing our work from Performance Test 2, we kept updating the AEV Project Portfolio, fixing the Arduino code, running more tests, and recording data for those tests. We completed the Arduino code, meaning the AEV covers the entire Jurassic Park scenario. Hermes is shown below.
Figure 1: Hermes
Hermes featured the use of two motors and two blade propellers, as well as an arm to properly distribute weight across the base of the AEV. In addition, the battery was placed next to the arm of the AEV, held by strategically positioned zip ties. The rudder was also designed to be on the bottom to hold the two propellers, similar to The Excalibur.
The data for our test run is shown below.
Figure 2: Propulsion Efficiency and Power Input vs. Distance for Hermes
Figure 1 above shows the propulsion efficiency and power input vs. time for Hermes. From the figure, it can be seen that the propulsion efficiency reached peaks of close to 14.50% but leveled off throughout most of the test right around 12.25%. The power supplied to the AEV was found to be in the range of 4-6 Watts when bursts of speed were necessary in the program.
Figure 3: Propulsion Efficiency and Power Input vs. Time for Hermes
Like Prototype 1 from Performance Test 1 and The Excalibur from Performance Test 2, propulsion efficiency and power input vs. time and distance were found to be very similar to each other, with relatively the same values in each graph.
Hermes paralleled the propulsion efficiencies of The Excalibur, but with a secondary motor, was able to move in the opposite direction as well. The use of a second motor in the opposite direction allowed the the AEV to operate in its most efficient orientation regardless of the direction it was traveling.
The Arduino Code for Hermes is shown below.
Table 1: Phases and Corresponding Arduino Code, Distance Traveled, Time, and Total Energy
| Phase | Arduino Code | Distance Traveled (m) | Time (s) | Total Energy (J) |
| 1 | reverse(4);// Run motors for burst of speed to first gatecelerate(1,0,40,3);motorSpeed(1,39); goFor(2);brake(4); | 0-1.2 | 0-5 | 16.8980 |
| 2 | //break at middlegoToAbsolutePosition(330);motorSpeed(2,30);goFor(1);//stopp at gatebrake(4);goFor(7); | 1.2-4.5 | 5-10 | 5.0238 |
| 3 | //stopp at gatebrake(4);goFor(7); | 4.5-4.8 | 10-17 | 0.7481 |
| 4 | //proceed to endcelerate(1,0,40,3);motorSpeed(1,33);goFor(2); | 4.8-8.2 | 17-25 | 14.6584 |
| 5 | brake(4);//Pause at endgoToAbsolutePosition(714);brake(4);goFor(7); | 8.2-9.8 | 25-35 | 1.2528 |
| 6 | //proceed to middle gatecelerate(2,0,40,4);motorSpeed(2,35);goFor(6);brake(4); | 9.8-14.2 | 35-48 | 45.7878 |
| 7 | //brake at middle gategoToAbsolutePosition(450);motorSpeed(1,40);goFor(1);//pause at middle gatebrake(4);goFor(7); | 14.2-17.6 | 48-55 | 0.7041 |
| 8 | //return to home celerate(2,0,40,4);motorSpeed(2,40);goFor(1);motorSpeed(2,35);goFor(6); | 17.6-19.7 | 55-70 | 46.1369 |
In the table above, the power output and distance travelled is outlined with regards to the different phases. These phases are determined but the Arduino code that programs the different phases. As seen above, the energy output is the largest in the phases where the motors are programmed to propel the vehicle forwards. It was discovered during testing, that it would be advantageous to use the motor as little as possible. As seen in the table, in phases that did not code for the direct use of the motor, there was a significantly greater decrease in power output.
The strategy for creating the code that operated the AEV resulted in an effective AEV. The AEV was run and then errors that needed to be improved were documented. The group proceeded to discuss ways in which the Arduino code could be improved to better complete the tasks assigned and make this process more efficient. For example, if the AEV came too fast into a checkpoint the code would be adjusted to come in slower. This would allow for more energy to be saved. One specific way the AEV was able to conserve energy was during the approach to the caboose. Initially the code used a reverse motor to break the AEV. The code was programmed to allow the AEV to approach the caboose slower and did not need to use any extra energy to stop itself. This was helpful in making the AEV use less energy.
Occasionally the AEV magnet would disconnect from the caboose as it was rounding the first curve after being picked up. This was solved by a hardware adjustment not by a code adjustment. The screw that connected the bracket to the AEV was loosened slightly. this allowed for it to turn slightly as the AEV turned and this problem did not occur again.
The group concluded that the best way to create an efficient code was to decrease the amount of time the motors were on. The power provided to the motor occasionally was increased to compensate for this but it ultimately ended up being more efficient. This strategy necessitated the usage of coasting. This played into the strengths of the AEV because there is minimal friction between the wheels and the track on which the AEV is connected.
Performance Test 3 combined the ideas of Performance Test 1 and Performance Test 2. In Performance Test 1, Prototype 1 had a high energy efficiency, but also a high combined cost. In Performance Test 2, The Excalibur had a low energy efficiency, but also a low combined cost, the exact opposite of Prototype 1. Each prototype had a good characteristic and a bad characteristic. The goal of Performance Test 3 is to build a prototype that highlights both of Prototype 1 and The Excalibur’s strengths and discards their flaws. This resulted in Hermes, a prototype that inherited a low cost with a high efficiency. Hermes borrowed the ideas of The Excalibur’s low cost model, and implemented a secondary motor to match Prototype 1’s efficiency.
Through experience and multiple tests in Performance Test 1, Performance Test 2, and Performance Test 3, an efficient and complete code had been programmed to complete the entire Jurassic Park scenario. Therefore, the data shown in the graphs cover the entire track and the return trip. The efficiency was measured to be 14.5% at its highest and 12.25% most of the track. Although The Excalibur had an efficiency of 15% at its highest, the graph shown covers the entire track, unlike The Excalibur which only covered half of the track. Using this logic, Hermes can be labeled as having a high efficiency. In adding an additional motor, propeller, count sensor, and sensor wire costs $14.45. This makes Hermes have a total combined cost of $156.95, which is still cheaper in comparison to Prototype 1, which costs $186.86.
After Performance Test 3, there is one final test, called Performance Test 4. In this last test, there are some improvements that could be made. Even though the Arduino code covers the entire Jurassic Park scenario now, there still exists some minor discrepancies that could be refined. These are usually stopping an inch too soon or late before a gate, or the magnet sometimes not attaching to the caboose. Another improvement that could be made is improving the efficiency. It is at an acceptable value at present, but could be made better.