In this blogpost, we are looking back and reflecting upon the entire project. What did we like and what did we dislike, and why? What tasks went great and where was room for improvement? This reflection is divided into the following sections:
- Communication with supervisors
- The framework
We are starting off this reflection on a very positive note: our teamwork. We can all agree that we are very satisfied with the way we worked together as a team during this project. Agreements were complied with, tasks were divided equally and we have honestly had no significant disagreements. Because we were constantly communicating and updating our planning, potential problems were easily prevented.
Furthermore, we made good use of our individual talents and we learned from each other by sharing our expertise. All of this was possible because we were all very motivated. The fact that we were so motivated, communicating so well and the fact that we were working together at the faculty every single day, made us a great team.
Communication with our supervisors
As for our communication with our supervisors, there was some room for improvement. Our main mistake was that we were seemingly ‘going with the flow’ while we actually did not know what we were doing was right. We let many questions be unanswered, which caused stress at some moments during the project.
The main thing we could have done better, is communicating with the subject coordinator: Jouke. After the main introduction to our project, we still were not completely sure if our assignment suited the main goals of the project. We were worried that we would do the assignment correctly, but get a bad grade because the assignment did not fit the project. In hindsight, this did not turn out to be a problem. However, we could have saved ourselves some stress by immediately going to Jouke with our concerns. The first time we actually spoke with him, was in the last week of the project, which would have been way too late had the assigment been insufficient.
The movement of light
In this section, we focus on our goal to create an eye tracking setup in which we could change the light exposure of the painting during an eye tracking test. In the first week, we were given an Arduino driver and an Arduino stepper motor. We wanted to use these to move a light source in a consistent way.
Our first mistake was to assume that the material we were given, was working. It turned out that both the driver and the stepper motor were broken. However, we only found out about this the day before our pilot. This meant we quickly had to think of a plan B to make the light source movable. We are actually proud of how quickly we came up with a working plan B. However, looking back, we should have thought of another way to move the light way earlier. That was our second mistake. Keeping another option in mind for moving the light in an earlier stadium, would have saved us some stress.
Building the framework went well and quickly and the final frameworks are simple, yet effective. However, there are things we could have done better.
The designs we made for the frameworks, were based on wooden beams. We never used wooden beams though. Our main supervisor, Willemijn, told us we could use some spare aluminium frames from the Applied Labs. We used these frames, but we did not really change our frame designs in a beneficial way. Looking back, we could have used the properties of the aluminium frames to create better, modular frames. By making the frames modular, the research setup could have been suitable for way more situations. This does not mean our current frames are not good enough, it just means they could have been better.
Our pilot had a rough start, because we could not use our regular beamer. Therefore, we had to adapt our entire setup to a different beamer in the morning. Had we made our framework more modular, situations like these could have been solved quicker.
The rest of the pilot went fairly well. However, we made the mistake of asking parents with kids as test subjects in the morning, and only students in the afternoon. This meant we had a great age variety for the real painting, and no age variety at all for the 3D printed painting. Not necessarily a big problem for the purpose of our pilot, but it is a bit sloppy.
As for the survey, we were too focussed on the perceived realisticness of the paintings and its causes. When analyzing our results, we realized that we should have focussed more on the precision of our eye tracking setup instead. Willemijns goal is to analyze the perceived realisticness, while our goal was to create a working, precise setup for her. In our enthusiasm we had lost track of our goal.
As for the iterations we made, we are very satisfied. During the project, we stayed critical towards our work. Though we could not improve everything there was to improve due to the short timeframe we had to work in, our final eye tracking setup has become way better than our first. Our final setup is more precise, easier in use and more comfortable than our very first setup.
All in all, we are very happy with the final product, which includes the prototype for the eye tracking setup and our blog. Though there is room for improvement, we feel like we gave it our best. You live and you learn.