Skip to Content

Changes for page Rick

Last modified by Rick Dekker on 2023/04/11 13:38
From version 7.1
edited by Rick Dekker
on 2023/04/07 12:23
Change comment: There is no comment for this version
To version 8.1
edited by Rick Dekker
on 2023/04/11 13:38
Change comment: There is no comment for this version

Summary

Details

Page properties
Content
... ... @@ -1,4 +2,3 @@
1 -(% style="line-height:1.38" %)
2 2  (% style="color:#000000; font-family:Arial; font-size:11pt; font-style:normal; font-variant:normal; font-weight:400; text-decoration:none; white-space:pre-wrap" %)**Week 1: **(%%)
3 3  (% style="color:#000000; font-family:Arial; font-size:11pt; font-style:normal; font-variant:normal; font-weight:400; text-decoration:none; white-space:pre-wrap" %)Learned the context of the use case. Started working on the quick start and pretty quickly settled on a concept. Our group is now going to make a pet-like robot as a social companion. The quick start lecture was super useful to get started. I had missed the first lecture, so I missed a bit of context, but the quick start lecture got me going pretty soon.(%%)
4 4  \\(% style="color:#000000; font-family:Arial; font-size:11pt; font-style:normal; font-variant:normal; font-weight:400; text-decoration:none; white-space:pre-wrap" %)**Week 2:**(%%)
... ... @@ -10,4 +10,8 @@
10 10  \\**Week 5:**
11 11  While Mohamed and Clemente improved the Wiki, Mees and I continued working on the robot code. After a discussion with the teacher we decided to switch our use case from "going on a walk with the patient and following them" to "interacting with the patient". "following the patient" was mostly interesting from a computer science perspective and less from a social-cognitive perspective, while "interacting with the patient", really challenged us to implement social cues and design a social robot. We first had the idea to implement "petting on the head" and "petting on the back" as two functions, but noticed that Miro would sometimes register its' own movement as a "pet on the back". When we noticed this we decided to switch to "petting on the head" and "clapping" as ways to interact with Miro. Sometimes however Miro would register the sound of its' own movement as a clap. To fix this we had to turn down the reactivity of Dogg0 a lot and basically made it wait to listen for a clap a few seconds after it had stopped moving.
12 12  \\**Week 6:**
13 -We spent a lot of time figuring out how to make Miro work on the floor. Previously we had done the experiment on a table and that worked well. When we put Miro on the floor it stopped moving forward however. It turned out that Miro saw a dark surface as a cliff and that it would not work on the black floor. Eventually we found a grey floor were one of the available Miro's would work. This week I also spent a lot of time finding a metric that reflected how much participants would trust Dogg0 and the interaction. Eventually I found an interesting paper about it and could include this in our questionnaire. This week we also did our experiment. We found 10 participants who were willing to do the evaluation and made them fill in the questionnaire. I started analysing the results and started working on the presentation. Our experiment was not very well designed, we did not have a null-experiment and our mood score was inspired by acedemic papers, but not the same. I feel like with a better set up experiment, our project might have been a lot more rewarding. I was also hoping to have had more inspiring guest lectures like the one we got from Rebecca Schaefer by this time, I would have loved to have some more practical expert knowledge from different disciplines.
12 +We spent a lot of time figuring out how to make Miro work on the floor. Previously we had done the experiment on a table and that worked well. When we put Miro on the floor it stopped moving forward however. It turned out that Miro saw a dark surface as a cliff and that it would not work on the black floor. Eventually we found a grey floor were one of the available Miro's would work. This week I also spent a lot of time finding a metric that reflected how much participants would trust Dogg0 and the interaction. Eventually I found an interesting paper about it and could include this in our questionnaire.
13 +\\**Week 7**
14 +This week we did our experiment. We found 10 participants who were willing to do the evaluation and made them fill in the questionnaire. I started analysing the results and started working on the presentation. Our experiment was not very well designed, we did not have a null-experiment and our mood score was inspired by acedemic papers, but not the same. I feel like with a better set up experiment, our project might have been a lot more rewarding. I was also hoping to have had more inspiring guest lectures like the one we got from Rebecca Schaefer by this time, I would have loved to have some more practical expert knowledge from different disciplines.
15 +\\**Week 8**
16 +This week we did our final presentation, which I feel we did acceptably. It was good to have a final feedback moment before finishing up the XWiki. The rest of this week we spent on improving the XWiki, polishing the details, better analysing our results and filling in missing pieces.