Clara Stiller
Week 1:
The first week of lectures I missed due to a covid infection. Therefore I followed along with the slides and got in touch with my group members.
In the group chat, we started planning our project based on Cesar's idea. It took me some time to find out, what the assessment will be about, and got familiar with the XWiki page.
Week 2:
Still missing in presence due to quarantine, but I could join virtually for the exercise session on Tuesday.
We divided the work in the wiki, I worked on Stakeholders. For better understanding, I went through the provided SCE documentation (Guide, FAQ, Glossary)
Week 3:
I started to search for literature about similar projects to ours. The information I found, I used to write the introduction and section about how we used music in our project.
Also, we decided on focusing on use cases 1,2 and 4. Use case 3 ('Guide user when going out') we left out because it would be out of scope for our project. Together we worked on the Personas and created Storyboards.
When creating the storyboards we recognized some problems and challenges that might come up during implementation. For example, it became clear that we can't provide dialogues for each reason to leave because there are too many possibilities. Another problem that might occur is the knowledge the robot has to have about each PwD. To react appropriately and personally, the robot has to have a broad knowledge about the daily routine, favorite activities and stage of dementia of each PwD. Since it would be too time-consuming to type in all this information, we decided to provide one general dialogue that suits to most situation.
Week 4:
During this week everyone worked on claims and functionalities. Also we fill in the evaluation part and planed our study.
Together with Cesar and Xin, I prepared the presentation for week 5.
Week 5:
On Tuesday we gave the presentation. I revised the evaluation part with the feedback we got after presenting our project. I changed the research questions to focus more on the dialogues, comparing the mood of PwD and how effective the music is.
Everyone of us started to get familiar with Choregraphe, to start with first tests on pepper in the next week.
Week 6:
We finally tested the pre-programmed dialogue flow with pepper, which needed a lot of debugging and fine-tuning. Therefore we had several sessions in the lab where we worked together until the first prototype was almost done.
Week 7:
Both prototype implementations were done on Tuesday as well as the questionnaire, so we could start with the evaluation on Wednesday and Thursday. Again, we divided our workload: some of us where doing the evaluation, some started to prepare the presentation or edit some pages at the wiki.
I also participated in 3 other evaluations from different groups. It was interesting to see what they had worked on and how well their prototypes worked.