Choosing driving simulator for my experiment.
Liza Yusoff, PhD
A storyteller | Data enthusiast who fancy statistics and machine learning
I was attached to Institut Image, Chalon-sur-Sa?ne, France, during my PhD study. Many people asked why and how. It is too complex, with multifaceted reasons, to write in a post. So if we get the chance to catch up in person, I'll share. But what I want to share in this post is my experience in choosing a driving simulator for my experiments.
Institut Image had many virtual reality setups, but the one relevant to my study required a setup that included a driving simulator in a virtual reality environment. There were four choices:
The plan was to use SAAM for my experiment. Of course, I was ecstatic – playing with real toys is basically like having my dreams come true. But nothing worth having comes easy – and most of the time, what I really wanted was never the best for me.
Juggling between making sense of my research from multidisciplinary perspectives (signal processing, neuroscience, psychology, virtual reality), designing experiments that were feasible to answer the research questions (governed by the title, that was immutable due to funding), and utilizing only the equipment available at the lab (nobody wanted to pay for new toys). I also worried about whether I could recruit enough subjects to provide sufficient power for the analysed data and only God knows what else. It makes sense why I’ve become an overthinking creature to the max, extremely introverted – I only got one shot at this, I couldn’t afford to make mistakes.
I started off with a little side-project left by the previous post-doc (because he got a better offer) – Vicariance project. It was a collaboration with Professor Jean-Michel Boucheix, Professeur de Psychologie Cognitive, Université de Bourgogne and his master’s students, Maud and Lea. They already had the objectives and some ideas on what data to collect (eye tracking and driving measures). They needed someone to setup the experiment, workflow, data acquisition and work on the technical stuff (only God knows, it was way more complicated than what was written here) – while Maud & Lea would recruit the subjects and conduct the experiments. Logitech driving simulator and projector was enough for this project. Here is the link to a video that I sent to eye tracker manufacturer to troubleshoot an issue: https://youtu.be/uXmVLhzcNac
After Vicariance project, I carried on conducting another experiment (because the EEG device just arrived from Malaysia – by this time, 8 months already passed) and at the same time, evaluating SAAM and ECO2 to see if they were suitable for my driving distraction experiments. Both were a no-go. There were too many problems to tackle and too many variables.
Finally, we resorted to CAVE + SI2M. Even this one had many issues, but it was still doable (since I had experience with Robocon in UTP, and I have an EE degree, somehow these electronics things make sense to me) so I had faith. I remembered my SV’s face turned blue when I refused to do the experiment at Renault’s lab in Paris (after explaining the challenges that might occurred) and decided to work on CAVE & SI2M. At this moment, I had excellent support from everyone, for which I am eternally grateful. Suddenly, everything worked in my favour.
I spent about a month working on setting CAVE + SI2M and two weeks designing and implementing everything in accordance with my research, - dry run, pilot run, testing run, you name it. Here is an example of video that I made for manufacturer to troubleshoot issues in CAVE: https://youtu.be/ZLT6VsJMaX0 .
领英推荐
Since I had to conduct the experiment solo, managing only God knows how much equipment, I was petrified. And by then, I’m left with four months and two weeks – in which I need to recruit and conduct the experiment.
Everything went well, until……… yep, nothing ever worked out my way. Someone messed up the setting in between the driving simulator and the software. Now the driving data can’t be recorded, nor can the driving simulator be controlled by the software. I think that was the first time, I couldn’t control myself and became the monster that I was – I shouted at the poor boy who made the mistake. It’s one of the moments that I regret the most; it felt like I’ve climbed Everest only to find myself at the bottom again.
I sort of lost myself after that. I couldn’t afford to fork out any more of my savings (I had already used a whole lot of them) to stay there; especially to stay longer just to collect data (that I’m not sure if I could by then) when it was not even my fault to begin with. But I couldn’t afford to entertain any of my feelings because I only had like 2/3 weeks left before returning to Malaysia – I bottled them up, never process it, I think that’s why I felt a little “cuckoo” when I returned home – I’m really sorry to those who asked why I’ve changed and got a lash from me, I mean no harm – I was hurt and broken beyond words.
So, I looked at the data that I’ve already collected as well as Vicariance data, and I found interesting result in Vicariance data that was not discovered by the master’s students. Told my SV about it, and somehow, we ended up with a meeting with Professor Alain Berthoz, Professor College de France along with my SV (Professor Frederic Merienne), Professor Jean-Michel Boucheix and Professor Andras Kemeny from Renault.
During the meeting, I showed an eye tracking video of one the participant with its mapping. Prof Berthoz was so consumed with it (I felt like watching myself being curious), and he pointed out something about the way the eyes were focusing on the edge of the road while driving – this trait had been reported and granted Nobel prize or publication in Nature (I don’t remember – honestly, I don’t really care). But what fascinated me was that the work I had done was of such quality that he could point out a trait that is known to him but not to me or everyone else in the meeting. Maybe Prof Boucheix knows about it – he did sent email the day before when he saw the video, stating that the result was very good. Finally, some real validation.
I felt like this meeting was a gift, a way for Allah to console my heart after what had happened. I was over the moon. I knew I did something right, and I’m glad I received that validation. Alhamdulillah, that’s what I’ve been looking for, and since then, I want nothing else of this world. I am very content, and my perspectives have changed in many ways. I now know that whatever hardships I need to face are God's gifts.
That's how choosing a driving simulator for my experiment has impacted my life.