Summary
For our research, we decided to pursue a mix of primary and secondary sources to gain a holistic view of our space. We started with an extensive literature review before launching two rounds of surveys. After the surveys, we followed up with a few respondents for interviews. Our goal was to identify problem areas in the current long-distance relationship dynamic that robotics and interaction design could help ameliorate.
Literature Review
Our literature review focused on robots providing physical affection through embodiment. Here, we summarize a few articles that may be relevant to our project.
Robots in Society (S. Rainey, “Friends, robots, citizens?,” ACM SIGCAS Computers and Society, vol. 45, no. 3, pp. 225–233, May 2016.)
- Searle sees robots as inauthentic motivation to the most human-like behavior with social conventions only as a list of rules and not truly ‘understanding’ human civility
- Aristotle believed that we only give true citizenship to those we believe are equals
- Kant stated that citizens must participate in communal sense
- Habermass said that a robot will not be seen as a social equal as long as they are not perceived to be free
Robots in Love with Humans (H. A. Samani and A. D. Cheok, “Probability of love between robots and humans,” 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2010. H. A. Samani and E. Saadatian, “A Multidisciplinary Artificial Intelligence Model of an Affective Robot,” International Journal of Advanced Robotic Systems, vol. 9, no. 1, Jun. 2012.)
- Lovotic’s Love Assembly State Module has 13 parameters which are often highly coupled and implemented in a Bayesian Network
- Affection State Module has 3 dimensions of motivation, activation, and the rate of change of the other two
- Artificial Endocrine uses the robot’s physical and network situation as well as audio and touch stimuli to simulate four emotional and six biological hormones in a Dynamic Bayesian Network
- Mixed in an ANN and used to control robot’s tilt, vibration, color, motion, rotation and height
Humans in Love with Robots (H. Samani, “The evaluation of affection in human-robot interaction,” Kybernetes, vol. 45, no. 8, pp. 1257–1272, May 2016.)
- Evaluated on six love styles (Eros, Ludos, Storge, Pragma, Mania, Agape)
- Got positive results with users perceiving the robot loving them slightly more than them loving the robot
Ethics (J. P. Sullins, “Robots, Love, and Sex: The Ethics of Building a Love Machine,” IEEE Transactions on Affective Computing, vol. 3, no. 4, pp. 398–409, 2012.)
- “When your robotic lover tells you that it loves you, should you believe it?”
- Robot relationship could become pathologized as people become attached to and addicted to a sexual/romantic/friend partner who is selflessly devoted to the owner
- Most important thing in sex/romance is reciprocation and how this can be simulated but not reliably created
- Evolutionarily humans are prone to assume love as a baseline and become attached even when reciprocation is doubtful or completely lacking
- This can be unethical since it takes advantage of evolutionary weakness to promote a fundamental untruth
- Love as a philosophical construct of making us morally better and are we morally better for loving a machine?
- There are two strategies:
- Variance. Make the love robot empathic but obviously a machine so that users are less likely to view them as long term partners but instead as enhancing our ability to love humans
- Mimesis. Make the robot seem as human as possible to enhance our love for them
Making It Happen (M. Jindai, S. Ota, and T. Sasaki, “A Hug Behavior Generation Model Based on Analyses of Human Behaviors for Hug Robot System,” MATEC Web of Conferences, vol. 82, 2016. H. Cramer, N. Kemper, A. Amin, B. Wielinga, and V. Evers, “‘Give me a hug’: the effects of touch and autonomy on people’s responses to embodied social agents,” Computer Animation and Virtual Worlds, vol. 20, no. 2-3, pp. 437–445, 2009. E. Y. Zhang and A. D. Cheok, “Forming Intimate Human-Robot Relationships Through A Kissing Machine,” Proceedings of the Fourth International Conference on Human Agent Interaction – HAI ’16, 2016.)
- Hug model based on angular velocities, pauses, and audio analysis
- A robot that engages in touch is trusted more when the robot behaves in an otherwise assertive manner than if it does not touch
- Robot is less trusted if it does so and behaves in an otherwise passive manner
- Kissenger combines mixed reality and lovotics
Surveys
Our surveys focussed on communication and the problems that people experience in long-distance relationships.
Google Form Survey Summary
- Current relationship length (~⅓ >2 years)
- Most relationships were temporarily long-distance (only 1 respondent indicated that they were in a permanent long-distance relationship)
- Primary means of communication were text messages & video chat
- Frequency of visits most commonly ranged from one month to a couple of months apart
- Open-ended questions revealed insight on the major challenges in long-distance relationships, which described:
- Arguments
- Risk of breakup
- Remote intimate activity
- Most found technology useful in maintaining their relationship
Mechanical Turk Survey
When we launched this survey, we specifically asked for participants who were in a long-distance relationship.
We found that
- Most were in their first long distance relationship
- Most relationships were between 6-24 months
- All the couples who responded were heterosexual
- Most couples were not worried about the distance
- Most couples saw each other monthly
Couples had a few ways of showing affection virtually:
- People used emojis or would say “I love you” via text
- Surprisingly, no one mentioned kissing screens, blowing kisses, or doing fake hugs on screen
There were certain physical activities that the participants missed as they could not do them in long-distance relationships. These included:
- Laying in bed together
- Kissing
- Cuddling
- Going to the gym
Interviews
Our interviews revealed interesting patterns in intimacy and communication.
Spectrum of Communications
- Mediums with low bandwidth (Snapchat, texting) used frequently and for short durations of time
- Mediums with high bandwidth (video calls) used sparsely and for longer periods of time
Uncomfortable Eye Contact in Video Calls
- Current video conferencing technologies present a barrier to establishing intimate eye contact
- Eye contact is a deeply intimate form of communication that requires both partners to see and be seen
- Participant cannot simultaneously look at the screen to see their partner’s eyes at the camera to let their partner see their own eyes (to compensate, one interviewee described a method of purposefully switching gaze between their partner and their camera)
Opportunities for Spontaneity
- Desire for more opportunities to engage in spontaneously intimate acts
- Intimacy is a continuous conversation that takes place over months and years, not minutes or hours
- For example, Snapchat “is a little part of my day that doesn’t always require a conversation [through texts]” in contrast to texting, which requires commitment to initiate and maintain
Benefits of Long-Distance
- Provides opportunity through adversity
- “It forces you to reevaluate your relationship.”
- “If you can get through this, you can get through anything.”
Insights
There are various actions that are missed in long-distance relationships. Some include physical contact, but others, such as eye contact or exercising together, simply require partners to be in the same room.
From this, we can conclude that our solution does not necessarily need to replicate the partner’s physical embodiment to be compelling. It could also focus on a non-physical intimate activity.
Intimate HRI 