Project home3. Human-Robot Collaboration
When it comes to human-robot collaboration and interaction, social factors governing human-human interaction apply equally [1]. This means that often human-human interaction models need to be considered and simulated to make the interactions feel natural and well-received. This means aspects such as gaze, tone, eye contact, and expressions need to be incorporated and these comprise essential parts of human-robot collaboration. Both communication channels and cues are also important [1]. This means that not only are the robot's own communicative abilities important when it comes to human-robot collaboration, but also the robot's ability to interpret and take in cues and communication given or displayed by the human counterpart in any way. This is central for our project also. Usually these communications are aided if the robot looks more humanoid, which for elderly care and for PwD's is very likely to be a central aspect, to make sure that interacting with the robot feels natural and does not stray away from regular human-human interaction.
Paper: Human-Robot Collaboration: A Literature Review and Augmented Reality Approach in Design.
https://journals.sagepub.com/doi/pdf/10.5772/5664
This is just a free search on the internet on the topic.
Important takeaways:
- Not only are the robots own communicative abilities important when it comes to human-robot collaboration, but also the robot's ability to interpret and take in the cues and communication given/displayed by the human. This is central for our project also. This is usually aided by robots looking more humanoid, which for elderly care or for PwD's is very likely to be very important, to make sure that interacting with the robot feels natural and does not stray too far away from looking at or interacting with a human.
- It is important to consider how exactly these cues and things will be taken in by the robot, what kind of sensors cameras etc will be responsible for it. This is also something we need to consider in our design, especially with the pepper robot who also has a tablet on the chest.
Regarding conducting studies in Human-Robot Collaboration, we have gathered the following information:
Firstly, focusing on the paper Conducting Studies in Human-Robot Interaction, many conclusions could be drawn when it comes to relevant evaluation methods in human-robot interaction studies. The relevant takeaways for our study regarded that it is essential to use three or more methods of evaluation to establish a study's validity and that additionally, it is also important to determine the appropriate sample size necessary to obtain significant results. [2] At this stage in our research, evaluation methods are really important to consider and this definitely aids us going forward.
Further, in order to conduct a successful HRI study the planning and design stages are really important. Here, we tried to consider what type of evaluation and what kind of study would fit our objective the best and we have considered either within-subjects design or between-subject design. We are still considering the time, the learning effect, and potential frustrations that might come if participants have too many similar interactions with the robot. The between-subjects design might be preferable in this regard, but the results could potentially be affected by the difference among the participants to a larger extent [2].
Another important aspect highlighted in this study is location and environment. It is mentioned that it is crucial that the study reflects the environment where the application is meant to be used and that conditions should be simulated accordingly and that this also ties into the choice of robot for the study [2]. In our case, it is important that we try to simulate the movement and walking of the robot together with the person. This means we need to consider a space that allows for some movement of the robot and participant together. Our goal is to simulate a walk in the garden and it is therefore important that we can mimic these conditions of a walk as closely as we can.
Furthermore, the exemplar study provided [2] was very useful to see a complete structure of a study, performed according to the guidelines. Conducting proper and successful studies is really important for the general field of HRI, since it is rather new, and these studies are always an important step in further normalizing and accepting robots as a part of society and as a helpful component in our lives [2], which is important to remember and hopefully something we can make a small contribution to through completing this course.
References
[1]
[2]