Skip to Content

Wiki source code of Simran - Self Reflection

Version 10.1 by Simran Kaur on 2023/04/11 20:00
Show last authors
1 === Week 1: Introduction to Socio-Cognitive Engineering ===
2
3 ==== Lecture: Introduction ====
4
5 The lecture introduced me to the standards of human-centered design and the concepts of socio-cognitive engineering. 
6 The ReJAM and PAL projects were explained along with videos. These demonstrated how a robot can be used as a social actor for stimulating social, cognitive, affective and physical processes through building a human-robot partnership. These projects served as incredible inspiration for building robotic intervention for people with special needs, such as children with Type I diabetes or elderly residents of Pieter Van Foreest suffering from dementia.
7
8 We were also introduced to dementia, and challenges faced by people with dementia as the domain for which we would build a hybrid intelligence system.
9
10 ====
11 Lab session: Storyboard, Quick Start ====
12
13 In the lab session, more details were given about the expectations from the project, and where to begin. 
14 My teammates and I started with an activity session in which we identified use cases associated with the challenges faced by people with dementia. My teammates and I used charts, sticker, and markers to create a storyboard around daily life situations faced by Georgina, the persona of our person with dementia. We also came up with personas and values after identifying other stakeholders, such as the formal caregiver, Eleana and the son of Georgina, Sam. We identified sundown syndrome, loss of interest in mealtime, and lack of interactions with her son Sam as some of the issues faced by her. 
15 The storyboard activity played a big role in getting our creativity fired up and after narrowing down on the mealtime use case, we came up with the idea of employing interactive storytelling in order to help Georgina build back her interest in mealtime, along with providing interaction opportunities with the people around her.
16 We put our thoughts and ideas into the Quick Start section in order to kickstart the design of the solution we wanted to build.
17 We used rest of the week to conduct background research on robotic intervention for dementia patients as well as interactive storytelling as a tool to help with it.
18
19
20 === Week 2: Music Interventions and Scenario-based Design ===
21
22 ==== Lecture: Cognition, music, memory and dementia ====
23
24 (% class="wikigeneratedid" %)
25 For this week's lecture, Professor Rebecca Schaefer gave a very interesting talk about how cognitive changes occur with dementia and how interventions can be designed and measured to enhance cognitive stimulation in a person with dementia. She also talked about the psychology of music and how music stimulates the brain. She further explained how music intervention can be designed for people with dementia and the benefits it can have.
26
27 (% class="wikigeneratedid" %)
28 While our use case does not directly involve music in the solution design, this lecture inspired me and my teammates to incorporate music in the interactive storytelling intervention that we were focusing on building. We discussed using pleasant background music during narration to provide a more immersive storytelling experience, along with playing the PwD's favourite songs while they ate. The cognitive stimulatory benefits of music therapy as discussed in the lecture provided for a solid foundation for this consideration.
29
30 ====
31 Lab session: Scenario-based Design, Foundation ====
32
33 The lab session revisited the steps of the SCE method, with a focus on the Foundation part and an emphasis on Inclusive Design. 
34 We also discussed in detail about our personas and their associated value stories, in order to inform our design decisions better. 
35 Then we were introduced to scenario-based design and its benefits in reflecting the current situation (problem scenario) and the envisioned future (design scenario). We used our personas and value stories to formulate the problem and design scenario which would then be formalized into our main Use Case.
36
37 We focused on defining our environment and stakeholders and further developing our personas and enhancing the problem and design scenarios for the rest of the week. Building the scenarios helped me to reflect on the reasons why we were building a human-robot interaction system in the first place, and allowed for visualizing how we would want the interactions to flow.
38
39
40 === Week 3: PAL project, Specification ===
41
42 ==== Lecture: PAL ====
43
44 In this week's lecture I learnt about how the PAL project used the SCE design cycle for building a hybrid diabetes management system for children with Type 1 diabetes. The lecture re-iterated the importance of stakeholder analysis and building value stories to inform our design. Further, I learnt about having a strong theoretical underpinning for our design through the example of how Collaborative Learning Theory is used for designing the PAL project. Along with this, we also briefly discussed that the choices of technologies for implementing our design should be intentional, personalized and affective.
45
46 ====
47 Lab session: Specification ====
48
49 In the lab session, we focused on how to extract a coherent set of use cases, functions, and claims which would serve as the Specification for our design from the Foundation built on operational demands, human factors, and technology. We focused on developing the objectives we wanted to design for, the requirements from our agent, and the personalized claims we would make. 
50 This week helped me understand how the design of a hybrid intelligence system is an incremental iterative process that grows by integrating technology into the situated practice via co-design by co-learning through building objectives, shared knowledge bases and learning by explanations and feedback.
51
52 For the rest of the week, my teammates and I focused on finalizing content for the midterm presentation.
53
54
55 === Week 4: Group Presentation - Midterm ===
56
57 (% class="wikigeneratedid" %)
58 My teammates presented the Foundation and Specification of our project. Our presentation was focused on detailing the environment, stakeholders, and persons for our system. We further dived into our problem and design scenarios, and our design plan for a robotic intervention for people with dementia through interactive storytelling, backed by scientific sources. The question/answer session helped provide a third-person view of how our design was perceived by others and how we could improve it. We got valuable feedback for the presentation, specially around designing our evaluation study for what was feasible for our scope.
59
60 (% class="wikigeneratedid" id="H" %)
61 I was also inspired a lot by attending the other groups' presentations and getting to know how they were approaching the design for their use cases.
62
63 We used the rest of the week to reflect on the midterm feedback and incorporate it into our design.
64
65
66 === Week 5: Design Specifications ===
67
68 ==== Lecture: Design Patterns, Design Rationale ====
69
70 (% class="wikigeneratedid" %)
71 This week we learnt about Team Design Patterns and Interaction Design Patterns which provided important concepts and tools for designing the interactions between actors for our system.
72
73 (% class="wikigeneratedid" %)
74 We explored the design questions for our system that would help address a general, broader issue. We also focused on supporting our design decisions with scientific theory and argumentation to fortify our design rationale.
75
76 (% class="wikigeneratedid" %)
77 We worked on building the IDP by exploring how the different actors in our scenario could and should interact with each other. This shaped the TDP for our main use case of the interactive storytelling session. Through this process, I learnt the importance of pattern-based design in identifying recurrent/generalizable problems within a context and defining structured descriptions for the solution to it.
78
79 ====
80 Lab session: Planning Evaluations ====
81
82 We also learnt about how evaluations should be planned for our prototype. The purpose of the evaluations is to assess the claims at the task level and check the usability of the interaction design. We were introduced to various approaches, like qualitative analysis to formulate hypothesis and quantitative analysis to test the hypothesis.
83
84 We explored the formative and summative evaluations that could be conducted for our system. We decided to focus on the summative evaluation to assess the overall effects of the system. Also, a formative evaluation would not have been feasible within the limited time.
85
86 For our measurements and metrics, we used the Godspeed questionnaire for a standardized assessment of our robotic agent. It address factors like being interactive, inert, animate, etc which were relevant for our use case. For the effects, we prioritized measuring the following: the mood of the patient post activity (subjective) and whether the patient finished their meal and in how much time (objective). However, we identified that due to the practical issue of not being able to test with actual dementia patients, the objective measures would not be collected accurately.
87
88 We wanted to host our questionnaires for online experimentation. For collecting the responses to our questionnaires, we decided to use the GDPR compliant Qualtrix survey tool.
89
90 This week was quite informative and learning intensive in terms of all the factors, practical and ethical, that need to be considered while designing an evaluation for a system.
91
92 === Week 6: Prototype Implementation, Pilot testing ===
93
94 (% class="wikigeneratedid" id="HLecture:Ontologies" %)
95 The focus of this week was implementing our prototype and testing it out within our group. We used Interactive Robots to program the Pepper robot with two pre-filled story templates - Picnic and Thanksgiving - for the testing of our Interactive Storytelling session use case. During this process, we built two versions of each story, non-interactive: one with simple narration and another, interactive: with inbuilt prompts to spark conversations. We planned to gauge the usefulness of the interaction design in our evaluations through comparison of the experimental interactive scenario with the control non-interactive scenario.
96
97 (% class="wikigeneratedid" %)
98 With implementation, we faced some challenges with the Pepper robot in using its tablet and getting it to recognize our speech and touch input. We were able to mitigate them by coming up with alternatives such as touch action on remote laptop screen through clicks, in order to preserve the flow of interactions. Working with the Pepper robot and being able to self-test the functionality we envisioned to build was quite interesting and further helped refine the design we had been building through our incrementally iterative design process.
99
100
101 === Week 7: ===
102
103 The focus of this week was conducting the evaluation for our prototype with participants and analyzing the results. We had prepared the participation consent form and the measurements questionnaire along with the prototype on the Pepper robot. Since we had limited time and a limited number of participants, we decided to conduct a within-study evaluation, wherein we would have each participant evaluate both the control and the experiment scenario. We also decided that for half of the total evaluation sessions, we would present the participant with the control scenario first, and for the other half, we would present the experimental scenario first. This strategy was employed in order to mitigate carry-over bias.
104
105 For each evaluation session, the participants first signed the consent form. Then they engaged in the first storytelling session, filled in the questionnaire, engaged in the second storytelling session, and filled in the questionnaire again. It was really interesting and informative to see how the participants responded to our interaction design, and how our system was perceived by them. Further, participating in other groups’ evaluation sessions also provided a broader view of the kind of effects our peers were trying to have with their design and what they considered worthwhile to measure.
106
107 With the evaluation completed, we analyzed the results through a statistical test to determine significant results. With this, our claims around improving the mood of the person with dementia through interactive storytelling were confirmed. The entire process taught me a lot about how to assess a system design in terms of the claims it makes for the effects it wants to achieve.
108
109 === Week 8: Group Presentation - Endterm ===
110
111 (% class="wikigeneratedid" id="H" %)
112 We compiled our project work and the results from the evaluation and I along with two of my teammates presented them to the class. We received interesting questions from our peers and the professors, which further helped us critically reflect on our design decisions. I also had an interesting discussion about designing the system in a way such that it minimises the possible negative effects that it could also have.
113
114 (% class="wikigeneratedid" %)
115 To conclude, designing a robotic intervention for people with dementia took us through the entire SCE process. It was a fruitful journey in which we learned by experimenting, analyzing and reflecting.