Andrei Stefan

Week 1

Attended the lecture on Monday and learned about concepts of Socio-Cognitive Engineering. Found a team to work on the project with and met together on Tuesday, when we created accounts for the XWiki and brainstormed some general ideas of what the robot needs to do, which robot would be most suitable for these tasks, and started writing some rough ideas of use cases together. We also created accounts on the Interactive Robotics platform, since we thought of working with Nao.
Individually, I studied the lecture slides again in detail and read this week's papers, in order to start writing the Human Factors and the Robotic Partner sections.

Week 2

Attended the lecture on Monday and learned about, music and cognition. It was interesting to see how big the differences regarding memory loss are between "regular" elders and people with dementia. We collectivaly started working on the Foundations tab of the Wiki, by spiltting into groups of two people. Xinqi and I worked on the Human Factors section. As I had already started writing the Robotic Partners part, sge suggested that I keep working on it, while she focused on wirting the Music and Cognition section.

Week 3

Missed the lecture on Monday, due to a meeting for a different course. I managed to catch up during the Tuesday lab, when we started working on the Specification tab of the Wiki. This week, my partners were Xinqi and Lang and we worked on the Use cases and the Claims sections. We spent most of the lab brainstorming which use cases would fit well for our choice of robotic partner and for the types of interactions between the patient and the robot that we envisioned (we decided to remove the tablet and only use the NAO robot). After a bit of struggle figuring out how the patient would communicate with NAO without a tablet for text input, Rembrandt found out that the robot can use speech recognition and thus hold conversations, which is the means of communication that we plan on using.

Week 4

Attended the lecture on Monday and learned about experiments and measurements. Together with the team, we divided the work for the week and started planning our experiments and thanks to Rembrandt who set up the timeslot, we had some time to play around with the NAO robot on Wednesday. While the navigation graph that Rembrandt had prepared didn't work as expected, we managed (after about three hours of struggle) to get the robot to understand speech. From there on, the only thing remaining was to implement the functionalities that we wanted. During the week, we also prepared the presentation for the following Monday and we decided that we would all present. Unfortunately, Rembrandt tested positive, so he could not attend, which meant we had to re-divide who would present what in a short meeting on Sunday.

Week 5

Presented on Monday and got feedback for how we should proceed with the project. Continued to develop the prototype, and by Friday we met up again to test it. Compared to last time, it went a bit better, as we didn't have to waste 3 hours trying to understand how to get the robot to recognize speech, but we still had some issues with it not wanting to run the program for no apparent reason. All in all, we managed to get a semi-final file to use for evaluation and filmed a short video showcasing the general flow of an interaction whith the robot (with some bugs still present, so it will at most serve as a bloopers video).

Week 6

Attended the lecture on Monday and learned about inclusivity. I also ended up writing the XWiki page about inclusivity, so having attended the lecture halped a lot. It wasn't a very eventful week. During the lab on Tuesday, we divided some of the pages to work on further. We didn't reserve a lab slot for this week, since some of us couldn't be there at the usual slot on Fridays, so we decided to rely on the robot to work fine during the evaluation (which, in retrospective, might not have been the brightest idea).

Week 7

Missed the lecture on Monday, since I had no selftests left and didn't want to risk getting everyone infected if I had Covid (which I didn't in the end, but better safe than sorry). On tuesday, the professor asked us about our progress with the evaluations, and, since we hadn't even started doing them, he offered to let us present on Tuesday so we woule have time to werp everything up, which was greatly appreciated. During the rest of the week, I picked new songs to replace the placeholders in the activities for the robot. While picking the songs, I wanted them to be royalty free, which came with the added benefit of them being only instrumental and somewhat similar to each other. I made use of this when designing the experimental plan too: since the music was similar, I also decided to order them differently that the informative paper we gave the participants, just to make it even harder to remember (so they didn;t have to pretend too much to have dementia). On Friday, I arrived at the robot lab in the morning, tested the robot to make sure everything was fine (and it was), but by the time my teammates arrived, the robot decided to not cooperate anymore (to the point where it wasn't recognizing words anymore). In a last-minute attempt to save the evaluation and not have to reschedule everyone, we made use of the tactile input on the robot's head to move further through the program, while I mimicked the replies he was supposed to give. Disregarding all of this, the evaluation went smoothly, and we had a total of 19 participants.

Week 8

Put together the presentation over the weekend and Monday and prepared to present on Tuesday. Presented on Tuesday and got feedback for what we can improve in the final week for the XWiki. We split up the remaining work for the pages, and all got to work. Looking back on everything, the course was enjoyable, even though working with robots turned out to be much more of a pain than any of us anticipated. Assuming we had more time, it would have been nice to get the robot to actually work and evaluate it properly, but I I feel like I still learned a lot despite this.