Cameras were a low bar for entry for the Air team; each student had a school-issued laptop and their own smartphones to use during the documentation process. When they saw “pictures” and “videos” mentioned on the initial graded rubrics, students immediately sought artifacts to take pictures of to meet requirements, without considering the purpose it served for their design process. This practice carried beyond the PDR assignment; during the complying and reacting phases, Samantha took general photos of the drone after repeatedly asking the team, “should I take a photo?” The number of photos piled up quickly. Although students in the team occasionally took photos or videos to share with friends, the team did not have a defined purpose for the photos they took until they reached a point in the building process where testing and iterating was daily practice. When Samantha took photos, she demonstrated an understanding that she should attend to the drone, broadly, but applied little to no interpretation to her capture, resulting in photos that were unhelpful in answering later questions, such as propeller placement.
On March 14th, London was programming the drone and removed the drone propellers to perform a tilt test, i.e. a test where someone holds the drone and tilts it sharply to hear the motors compensate for the loss of equilibrium (see Figure 14, below).
Figure 14 Performing the tilt test without propellers.
When the tilt test was over, London looked at the detached propellers and said, “Shoot. I don’t remember what propellers go where.” Attending to the need to see prior configurations, London asked if anyone had photos. “We didn’t write this down, did we?” Nick replied. The group discussed the problem, wishing they wrote it down but not having the information. “We didn’t intend to take the propellers off again,” London remarked. They did not want to have to watch the video again, because it wasn’t clear the first time. “I’m trying to think how we actually did it though,” Nick said, and added that he thought the configuration was different from the video somehow. Without photos to refer back to, there was nothing to interpret; in response, the team had to rewatch the original confusing instructional video, essentially redoing the same work they completed in a week prior and hoping that the configuration was the same. This was but one in a string of instances where students attended to the need to refer back to photos but had not taken photos to interpret. After that point, the team began to make a concerted effort to take a glut of photos and videos in an attempt to ensure that they had information about any configurations they might have used. It then became a struggle of locating the right photo to interpret.
On numerous occasions, when students finally found their photos or videos, they could not interpret from the photo or video what configuration the parts corresponded to. On April 2nd, the drone was once again disassembled post-crash. Nick attempted to figure out what direction the propellers and motors were spinning in during the crash flight to determine whether that was the cause of the crash. “Do we have a picture of what the drone looked like before we flew it?” Nick asked. He said he needed a top view. “Because if it’s the same, that wasn’t what caused the spinning,” Nick hypothesizes. He did not find the photo, despite Samantha’s organization of files in Google Docs. “I have one from before we put the propellers on,” Samantha explained. “I need the propellers on,” Nick insisted. “If you wait one second, Nick, I got you,” Billie offered.
Using whiteboards in conjunction with photos helped to improve the interpretability of photo and video data. At the first author’s suggestion, Nick wrote his hypotheses and diagrammed his understanding of propeller mechanics on the whiteboard. Eliot joined Nick to discuss how air flow would impact flight with his diagrammed configuration, and that the problem with the drone could be motor spin, propeller configuration, or both. The two decided what configuration the propellers should have been in and that they could do nothing more without comparing it to the drone’s crash configuration. When Billie found the photo, they were able to compare the configuration to their diagram and determine the problem. “They’re supposed to be diagonal,” Nick explains. They adjust the propeller design and Samantha writes the new configuration in their documentation, along with the link to the video that Nick found showing a different configuration. As the group came to make more use of the whiteboards, Nick in particular remarked that he found hunting for photos to be ineffective and unhelpful, and preferred to have information out on whiteboards where the team could see it.
By April 6th, during the iterating phase, the team had had enough of the confusion of unlabeled photos and videos. Students began taking photos more frequently and aimed their cameras to take more specific information, such as of parts of the drone rather than shots of students posing with drones. The team observed the new parts at their disposal included red and black propellers they could use to mark spin direction and they noticed their drone now had yellow zip-ties that corresponded to motor spin direction. They numbered the arms of their drone, like they had seen in the instructional video. From this point onward, any video recordings or photographs should contain enough information to determine the precise configuration and make it easier to determine any problems individual motors or propellers were experiencing. As shown in the photograph in Figure 15, next page, the students used colors, numbers, part names, and other labels in white board diagrams to explicitly map out their understanding of part positions and functionality and compare their drone setup to the instructional video they were using to build the drone.
Figure 15 The Air team’s labeled whiteboard diagram comparing their drone’s setup to that of an instructional video.
When they began testing, they found that even this was not always enough to ensure the capture of adequate information, for one very simple reason; the drone’s parts spun too fast and moved unpredictably during flight. Slow-motion capture on smartphone cameras saved the day. Students developed routines of capturing test flights and motion from different angles with a combination of slow-motion and regular video. In addition to the labels, this was enough information to interpret photos and videos, and to further ease the process, students immediately interpreted captured video to modify and update whiteboard diagrams, which proved to be the most convenient source of information for the team.