Yōkobo: A Robot to Strengthen Links Amongst Users with Non-Verbal Behaviours
, Pablo Osorio 1,†, Shohei Hagane 1,†, Corentin Aznar 1,2, Dora Garcin 1,2, Enrique Coronado 1,3, Dominique Deuff 4, Ioana Ocnarescu 2, Isabelle Milleville 5 and Gentiane Venture 1
Round 1
Reviewer 1 Report
This is a nice paper towards the concept of the robjects, human-centered robotics and the Society 5.0. This paper is also well written. Some improvements are suggested:
- Authors should consider including a MP3 video demonstration of this work, for example in a OS repository.
- The motor control description is missing.
- The software methods used in this work should be described with more detail.
- Some statistical mathematic procedures could be included.
- Is it possible to electrically charge the robot?
- What are the implications of this work, for example, for disabled people ?
- The conclusions should be also extended, considering the experimental data and the mentioned robot capabilities.
Author Response
This is a nice paper towards the concept of the robjects, human-centered robotics and the Society 5.0. This paper is also well written. Some improvements are suggested:
We thank the reviewer for their time and their comments that are improving the readability and comprehension of our work. We have all answered below and addressed them in the manuscript.
- Authors should consider including a MP3 video demonstration of this work, for example in a OS repository.
→ You can find videos on our lab Youtube channel: https://www.youtube.com/playlist?list=PLP8rVNZA2xVHpt_9j9SEAWuaaBFKJaomu (we will add more in future).
- The motor control description is missing.
→ The below paragraph has been added:
“The motors are controlled through the Python Dynamixcel SDK. The commands are given as a motor's position, then the Dynamixcel motors reach the commanded position using their inner PID controller.”
- The software methods used in this work should be described with more detail.
→ The following paragraph has been added:
“Since our approach follows a sequential loop of actions there was no need for parallelisation in the main states machines, the only subsystem running in parallel to MDL and MC is the data acquisition which constantly reads the sensors and publishes the data through NEP topics. This data is read whenever the FSM calls for it. The parallel process is in charge of acquiring the data and writing it to two specific buffers where the most up-to-date sensor values will reside. One buffer is uniquely given and access per FSM. Inside each FSM to avoid reading the buffer while the parallel program is writing, a semaphore method is implemented as a precaution. As proof of concept all programs are coded in Python, except for the HPEA which runs natively in C++.”
- Some statistical mathematic procedures could be included.
→ Our study focuses on the usability, initial reception from the users and readiness of Yokobo for future experiments. That is why we are not using deep statistical analysis with a larger group of participants nor acquiring a larger number of samples, but rather focus on UX methodologies.
- Is it possible to electrically charge the robot?
→ No, at the moment Yokobo has been created to be placed at the home entrance and not to be moved. Hence, a battery is not necessary. Technically it could be added if needed.
- What are the implications of this work, for example, for disabled people ?
→ Disabled people are not in the scope of Yokobo at the moment. Rather the implication with the young retired couples is the purpose of an experiment currently held at their home for several weeks. The result will be published in another paper. The scope of our paper here is to test the technical abilities of Yokobo before the real experiment.
- The conclusions should be also extended, considering the experimental data and the mentioned robot capabilities.
→ The conclusion has been extended especially with the new line of research implied by the experiments.
“Longer experimentations are ongoing with the target population to confirm some points of the findings. As mentioned in the discussion section, the novelty effect seems to have been overcome somehow. But with a longer experiment, over several weeks, we hope to observe what happens when a routine settles and the usage of the robot becomes seamless. Another point that can be pushed further is the feeling of the partner. With our two-week experiment, 40% of the participants could feel him/her, proving the capability of this concept. However, this number might seem still low, and one reason may be that building up this feeling may take time. Moreover, one of the limitations of our experiments is to use arbitrary couples, where the real couples in the ongoing experiments would have a stronger inter-connection; and it will be probably easier for them to feel their other half through Yokobo. The realised experiments gave us some clues about the validity of some of our concepts, as a first step.”
[...]
“Another way to improve the usability of Yokobo will be to add sounds linked to the movement of the robots. Even if the SUS gave us a good score, around the average, some improvement can still be done, especially about the recording process. However, since our goal is to be part of the slow technology movement with Yokobo, it is not necessary to have efficient and straightforward usability.”
Reviewer 2 Report
This manuscript describes the design and evaluation of a robotic object that shall facilitate communication between people through this object. The outlined concept is very interesting. However, the manuscript is vague on how this principle is implemented in the robotic object. Further, from the description alone, the stimuli used in the communication seem relatively weak and appear too abstract.
The research design looks fine, but there are shortcomings in the evaluation. The conclusions are generic and not further informed by the results of the evaluation. Please outline further, which of the results would suggest which type of further research.
Section 3 outlines the contribution of the manuscript. It is unusual to outline this in a separate short section. Please consider moving this to the end of the introduction.
There are some language issues that should be addressed, mostly terms that seem to be misleading. For instance, Line 22, the term "motion messages" is unclear.
Line 23: "while still serving as a key bowl" is incomprehensible at this point, as the purpose of the robotic object is not yet explained.
Line 24: the word robot is originally a Czech word that is assimilated into other languages. Suggestion: "... and the French word robot ..." --> "... and the French pronunciation of the word robot ...".
Line 26: "them" --> "sensor data". (unclear what "them" points to)
Line 28: "Unlike ...": This is unclear and needs to be discussed. There is a variety of household robots, such as vacuum cleaners or other automatic household appliances, that have an abstract shape and diverse means of communication. Household robots are not designed for communication, but might still contribute to facilitating communication between persons.
Line 33: remove "in" before [5]
Line 39: "A RA" --> "An RA" (due to pronunciation)
Line 49: incomprehensible "specificity to be a robject". The paragraph Lines 49-53 is unclear.
Line 74: Please add "relationship" in front of "H->R->H"
Line 80: Suggestion to move the reference after "Joeng et al."
Line 83: act --> acts
Line 89: remove "in" after "Campa"
Lines 107ff: Please move references directly after the authors' names.
Section 3: Please move to the end of the introduction and integrate. Note further, that one of this manuscript's main contributions consists of the evaluation of the robject, while design and implementation of the robject is more the means to facilitate the research presented in the manuscript.
Line 147: incomprehensible. What do you mean by "discrete object"?
Figure 3: the drawings are too tiny and are not readable. Please also use larger fonts and make use of the entire page width.
Line 268: Remove "." at the end of the line.
Line 269: unclear: How do you gather the usability metrics? Probably you mean gathering (quantitative) results using usability metrics yet to be described.
Line 298: unclear. What do you mean by "80% of the usability problems"?
In the experiments: How were the participants instructed? Please, consider outlining further details of the experiment protocol, how the participants were instructed, etc.
Line 319: incomprehensible: what is a "questionnaire amount"?
Line 322: why are Curiosity, Fearfulness, and Confusion capitalised?
Line 329: "is" -> "consists of"
Line 361: unclear: "agreed"
Figure 5: The drawing is unclear. The patterns in the diverse columns are not visible in the drawing. The "D" frames on the top are probably days, but this is rather unclear. Please use the entire width of the page.
Line 407: incomprehensible. What do you mean by "manners"?
Please also revisit Figure 6. It is further unclear how the interaction time can be interpreted. A short interaction time could be interpreted that the interaction is effective, while a long interaction time could be interpreted that the device is too complicated. However, this is probably not what you want to express. This needs further evaluation and discussion. ref also paragraph Lines 430-440.
Line 451: CPT? why superscript?
Line 449: Are there any observations that could support these claims?
Line 452: What do you mean by "with a slight standard deviation"?
Line 458ff: incomprehensible.
Line 468: Where do the semantic dimensions used in the manuscript have their origin? Please use a reference to the theories behind this. Please revisit the presentation and interpretation of the results. It is unclear on which scale the results are given. Also, the results in Table 1 are difficult to interpret. It seems that responses are on a Likert scale 1-5, where 1 means positive, 3 means neutral, 5 means negative.
Line 515: The Declaration of Helsinki is irrelevant for this research; it is relevant for medical research, and your research is not about healthcare or medical research.
In the questionnaires in the appendix, there are diverse language issues. E.g., "How many times did you interact ..." (s missing in times)
Table A2: incomprehensible: ...your partner existence
incomprehensible: "... was a Yokobo's behaviour ..." probably the "a" should be removed to mend this problem.
Table A3: several tenses are used here. Suggestion to use present tense only, i.e., "think", "find", remove "would", "feel", ...
Reference 21 is empty
Author Response
This manuscript describes the design and evaluation of a robotic object that shall facilitate communication between people through this object. The outlined concept is very interesting. However, the manuscript is vague on how this principle is implemented in the robotic object. Further, from the description alone, the stimuli used in the communication seem relatively weak and appear too abstract.
We thank the reviewer for their comments and for helping us improve the quality of our paper. Answers to the comments can be found below and highlighted in the manuscript.
The research design looks fine, but there are shortcomings in the evaluation. The conclusions are generic and not further informed by the results of the evaluation. Please outline further, which of the results would suggest which type of further research.
→ The conclusion has been extended especially with the new line of research implied by the experiments.
“Longer experimentations are ongoing with the target population to confirm some points of the findings. As mentioned in the discussion section, the novelty effect seems to have been overcome somehow. But with a longer experiment, over several weeks, we hope to observe what happens when a routine settles and the usage of the robot becomes seamless. Another point that can be pushed further is the feeling of the partner. With our two-week experiment, 40% of the participants could feel him/her, proving the capability of this concept. However, this number might seem still low, and one reason may be that building up this feeling may take time. Moreover, one of the limitations of our experiments is to use arbitrary couples, where the real couples in the ongoing experiments would have a stronger inter-connection; and it will be probably easier for them to feel their other half through Yokobo. The realised experiments gave us some clues about the validity of some of our concepts, as a first step.”
[...]
“Another way to improve the usability of Yokobo will be to add sounds linked to the movement of the robots. Even if the SUS gave us a good score, around the average, some improvement can still be done, especially about the recording process. However, since our goal is to be part of the slow technology movement with Yokobo, it is not necessary to have efficient and straightforward usability.”
Section 3 outlines the contribution of the manuscript. It is unusual to outline this in a separate short section. Please consider moving this to the end of the introduction.
→ As suggested the contribution section has been moved right after the introduction instead of after the related works.
There are some language issues that should be addressed, mostly terms that seem to be misleading. For instance, Line 22, the term "motion messages" is unclear.
→ We agree that this is a concept rather new and one of the contributions of our robot. A description has been added to explain the motion message. We hope the reviewer will now find it clearer.
Line 23: "while still serving as a key bowl" is incomprehensible at this point, as the purpose of the robotic object is not yet explained.
→ the sentence has been detailed to make it, we hope, more understandable
Line 24: the word robot is originally a Czech word that is assimilated into other languages. Suggestion: "... and the French word robot ..." --> "... and the French pronunciation of the word robot ...".
→ corrected according to the suggestion
Line 26: "them" --> "sensor data". (unclear what "them" points to)
→ corrected
Line 28: "Unlike ...": This is unclear and needs to be discussed. There is a variety of household robots, such as vacuum cleaners or other automatic household appliances, that have an abstract shape and diverse means of communication. Household robots are not designed for communication, but might still contribute to facilitating communication between persons.
→ changed “robot” to SR.
Line 33: remove "in" before [5]
→ corrected
Line 39: "A RA" --> "An RA" (due to pronunciation)
→ corrected
Line 49: incomprehensible "specificity to be a robject". The paragraph Lines 49-53 is unclear.
→ specificity has been changed to particularity. In addition, the paragraph has been changed to make it clearer. One sentence has been moved to another paragraph
Line 74: Please add "relationship" in front of "H->R->H"
→ corrected
Line 80: Suggestion to move the reference after "Joeng et al."
→ corrected
Line 83: act --> acts
→ corrected
Line 89: remove "in" after "Campa"
→ corrected
Lines 107ff: Please move references directly after the authors' names.
→ corrected
Section 3: Please move to the end of the introduction and integrate. Note further, that one of this manuscript's main contributions consists of the evaluation of the robject, while design and implementation of the robject is more the means to facilitate the research presented in the manuscript.
→ As suggested the contribution section has been moved right after the introduction instead of after the related works.
Line 147: incomprehensible. What do you mean by "discrete object"?
→ changed to discreet. By “discreet object”, we want to express the ability of Yokobo to fit in any home, without being too captivating or noisy, but rather being self-effacing.
Figure 3: the drawings are too tiny and are not readable. Please also use larger fonts and make use of the entire page width.
→ the figure has been enlarged
Line 268: Remove "." at the end of the line.
→ corrected
Line 269: unclear: How do you gather the usability metrics? Probably you mean gathering (quantitative) results using usability metrics yet to be described.
→ the sentence has been rephrased. We are gathering data regarding how the users interact with Yokobo, and by extension, we are able to get the usability metrics.
Line 298: unclear. What do you mean by "80% of the usability problems"?
→ The usability problems we are aiming to gather with our study are the negative phenomena that can frustrate the user during their interaction with the robot. A description of the usability problems has been added to the manuscript, and we are hoping it makes it clearer.
In the experiments: How were the participants instructed? Please, consider outlining further details of the experiment protocol, how the participants were instructed, etc.
→ A subsection about the instruction to participants has been added (§5.4)
Line 319: incomprehensible: what is a "questionnaire amount"?
→ changed to number of questionnaires
Line 322: why are Curiosity, Fearfulness, and Confusion capitalised?
→ corrected
Line 329: "is" -> "consists of"
→ corrected
Line 361: unclear: "agreed"
→ changed to pointed out
Figure 5: The drawing is unclear. The patterns in the diverse columns are not visible in the drawing. The "D" frames on the top are probably days, but this is rather unclear. Please use the entire width of the page.
→ The meaning of the letter D is indicated inside the caption of the figure. The figure has been enlarged, and the legend changed to be able to see the couples according to the participants.
Line 407: incomprehensible. What do you mean by "manners"?
→ We used manners according to this definition “behaviour that is considered to be polite in a particular society or culture”, but because it can bring some confusion, we added good before manners. We hope this is now clearer.
Please also revisit Figure 6. It is further unclear how the interaction time can be interpreted. A short interaction time could be interpreted that the interaction is effective, while a long interaction time could be interpreted that the device is too complicated. However, this is probably not what you want to express. This needs further evaluation and discussion. ref also paragraph Lines 430-440.
→ Sentences have been added at the end of the paragraph to clarify our expectations according to the interaction time.
“Due to the slow technology principles used for Yoboko, the more time the users spend interacting with the robot the better it is. We are looking for them to discover the functioning of the robot by themselves, and we are not looking for efficiency as it can be with usual robots or products.”
Line 451: CPT? why superscript?
→ corrected (Cpt refers to the name of the column in table 1)
Line 449: Are there any observations that could support these claims?
→ We rephrased the sentence to make it clear that it is one of our hypotheses (line 502).
Line 452: What do you mean by "with a slight standard deviation"?
→ we changed slight by a number
Line 458ff: incomprehensible.
→ the paragraph has been rephrased to make it more readable (line 514ff).
Line 468: Where do the semantic dimensions used in the manuscript have their origin? Please use a reference to the theories behind this. Please revisit the presentation and interpretation of the results. It is unclear on which scale the results are given. Also, the results in Table 1 are difficult to interpret. It seems that responses are on a Likert scale 1-5, where 1 means positive, 3 means neutral, 5 means negative.
→ We are using the semantic scale for the Table 1 data. Some other questions also use the Likert scale. To avoid ambiguity, we described the semantic scale in detail in section 5.3 along with the reasoning and literature support behind our words choice. We also indicated in the Appendix which scale is used for each question.
Line 515: The Declaration of Helsinki is irrelevant for this research; it is relevant for medical research, and your research is not about healthcare or medical research.
→ The ethics board we are applying to at our institution considers all experiments with humans as medical research. Hence, the declaration of Helsinki applies to us.
In the questionnaires in the appendix, there are diverse language issues. E.g., "How many times did you interact ..." (s missing in times)
→ corrected
Table A2: incomprehensible: ...your partner existence
→ changed to “existence of your partner”. Here, with these questions, we wanted to see if the participants feel the existence/presence of their partner through Yokobo. If they can see beyond Yokobo.
incomprehensible: "... was a Yokobo's behaviour ..." probably the "a" should be removed to mend this problem.
→ corrected with the suggestion
Table A3: several tenses are used here. Suggestion to use present tense only, i.e., "think", "find", remove "would", "feel", …
→ The questions used in this study come from the standard SUS questionnaire [https://doi.org/10.1080/10447310802205776]. We just replaced system/product with Yokobo.
Reference 21 is empty
→ error in the Bibtex file, corrected
Round 2
Reviewer 2 Report
The authors have addressed my comments sufficiently.
The content of Figures 5 and 6 could be better explained.
Institutional Review Board Statement: Please remove the reference to the H. Decl., as this does not apply to your research (although your board says otherwise). Please note: Point 6 of the H.Decs. states what type of research it applies to. Further, experiments that do not support the treatment of patients should not be performed (point 14). Thus, your research would not be supported by the H.Decl.
Author Response
The authors have addressed my comments sufficiently.
We thank the reviewer for their advice and comments to improve our article. We have answered the second round of comments below and addressed them in the manuscript as well.
NB: the modifications of the second round are marked in red in the manuscript with “minor” written superscript at the end of the sentence, in order to differentiate them from the first round.
The content of Figures 5 and 6 could be better explained.
→ We improved the descriptions of each figure in their respective caption. We hope it makes the figures more understandable.
Institutional Review Board Statement: Please remove the reference to the H. Decl., as this does not apply to your research (although your board says otherwise). Please note: Point 6 of the H.Decs. states what type of research it applies to. Further, experiments that do not support the treatment of patients should not be performed (point 14). Thus, your research would not be supported by the H.Decl.
→ Following the recommendation of the reviewer, we removed the reference to the Helsinki declaration.
