Skip to Content

Wiki source code of Personal Reflection - Liza

Last modified by Liza Wensink on 2023/04/07 15:59
Show last authors
1 **Personal reflection:**
2
3 I have thus far really enjoyed this course and the aspects it highlights. Human-computer interaction is the focus of my master's degree and it feels really great and interesting to be studying such a useful application of social robots that can really make a contribution to society in the future.
4
5 We decided to design a system that focuses on taking PwD outside for walks, in the garden of the care center. Ideally, the robot would be monitoring the surroundings during the walk and also making sure that the PwD does not go off track. This would ideally be a full system for the care center where the robot can log who has recently been on a walk and who hasn't out of the residents and then motivate the residents who seem to have a lower level of activity recently. This led us to focus on differences in effectiveness between goal-based motivation and emotion-based motivation when it comes to motivating PwD to go on walks. We started with the thought of wanting to only take PwD to parks but have limited our scope a bit due to feedback we have received during the design process. In this iterative process, we also came up with the logging system idea that could further extend the capabilities of our design. So far this process has given me many new insights and new knowledge. I have never previously worked with human-computer interaction from a motivation perspective and especially not when it comes to motivating PwD. I definitely think this is a very interesting topic to examine and dive into and something that could be relevant in the future and in projects where perhaps persuasiveness or a humanoid robot's ability to convince and motivate humans (and also specifically PwD) is relevant. I definitely think I will use the results and knowledge I gain from this project in the future.
6
7 Me and Maya did the first presentation together and it went well. We tried to be as clear as possible with our design and to present it in an interesting way for everyone. I had a good experience collaborating with Maya and we made a good team. We also received very useful feedback, both from our peers in the class and also from the professors and TAs. This was very useful for us to move forward!
8
9 We have also now (14/3-2023) had our first lab session with our selected Pepper robot, where I got to try to interact with the robot and piece together a small test program to run. I really enjoyed this lab session, it was honestly just really cool to see the robot and how it functions in real life and what it is capable of. I had first and foremost a really fun time with it and seeing that we were able to make it execute some commands we have planned. This has further peaked my interest in the project and for completing it, but also for potentially working with humanoid robots and their interactions with humans in the future. Overall, I am feeling quite inspired after this session.
10
11 (21/3-2023) Me, Demi and Maya went to test our new version of the prototype, which Demi had written using IR, with Hugo's schemes as a base. We discovered, however, that IR is not able to run on the Pepper robot anymore and that connecting to it seems to not be possible either. Earlier in the day we had spoken to the TAs who has advised us to maybe use Choregraphe instead since we also had trouble making the Pepper robot go walking through IR. Therefore, I attempted to rewrite the IR programs using Choregraphe before we arrived at the lab. Some parts of the program were able to run while others posed bigger problems, such as speech recognition. We struggled with getting this to work quite a bit and ended up deciding to book multiple other sessions later in the week to try and solve this before we dive into the evaluation.
12
13 (23/3-2023) Me and Demi had a lab session with the Pepper robot where we continued to attempt to solve the Speech Rec. issues mentioned previously. This involved searching for solutions on the internet, asking Ruud questions, and trying to also communicate with TAs from the course. Ultimately we spent a few hours on this but didn't end up finding a complete working solution. We did however understand that it has something to do with the ASR engine, which needs to be turned off and on again in the Speech Rec. process.
14
15 (24/3-2023) Maya and Hugo managed to get the Speech Rec. working and completed the programs for the different motivation types, which completed our prototype!
16
17 (27/3-2023) We met in the lab on this day at 13:00 before the Monday lecture in order to complete our pilot test, before it is time for the real evaluations, which we have scheduled for Wednesday with our participants. This turned out to be quite useful since we ran into some problems regarding speech recognition again. It wasn't quite working or recognizing the words as well as we wanted it to, which posed some problems with the functionality of the prototype. We ended up solving these, however, by editing our Choregraphe program a bit, which left us with a working prototype for Wednesday. We also made sure to run the pilot test like a real evaluation test complete with our orientation script, consent form, and questionnaire to make sure everything is in order for Wednesday.
18
19 (29/3-2023) The evaluation day ended up being quite long for us, we stayed in the lab from 09 until 17 and managed to do 16 evaluations in total. Based on the literature it is, of course, preferable to have about 15 users per group and we really tried to get as close to the number as we could but ended it with 8 participants per approach. This was sufficient for this study in particular, but not ideal. We also realized in the beginning while testing the robot before the evaluations started that we had missed handling some essential things when it comes to the "What do you think of the weather?" interaction. We decided to tweak this and handle some different cases of different replies the participants might give and after some testing, we got this to work for our tests. Overall the day was successful and I thought that we made a good team and kept to a well-structured way of evaluating.
20
21 Overall, something that I have really enjoyed with this course and learned a lot from is all the literature I have read regarding human-robot or human-computer interaction and collaboration. I have also read literature regarding the use of music and the use and technique of portraying emotions in robots, through for example colors or body language and how this may influence interaction. I also read up on the persuasiveness of conversational agents which is a really interesting topic in my opinion and not only does this all relate to my choice of study, but I also feel like it is highly relevant for the future. I am definitely interested in diving into these kinds of fields and topics further and I find the intersection of human behavior and psychology together with technology and robotics highly interesting. I also was really excited and inspired by working with the robot, I am looking forward to when I can do something similar, and while the design we worked out ended up being quite limited I was excited by the idea of working on it further and all kinds of different possibilities made me really interested and excited to work on a more refined and detailed version. I think it would be really satisfying to create a design that is really extensive and detailed and see it in action!
22
23 I also would like to give my compliments to the course, its TAs, and the professor. I really enjoyed this course and the topic of study is really interesting to me, I really enjoyed all the literature that was provided and the literature that was necessary to go into for this project. I also really enjoyed working with the robot and I think this course has really opened my eyes to what is possible and what I would like to do in the future. I think the TA's and professor's expertise within the field has been great to learn from. The only thing I would have wished to see is occasionally more clarity on what exactly is expected from certain assignments or steps in the process, but it was all still understandable and clear after asking questions.