There’s always something heart-wrenching when parents have to be far away from their children. Sometimes parents travel. Sometimes separated parents split time with the kids.
Maybe telepresence robots can help.
This article explores some of the potential benefits and challenges of using telepresence robots as a co-parenting tool. To my knowledge, there is only one research instance of this [pdf], so a lot of questions remain to be answered. What is clear is that there are reasons why co-parenting with telepresence robots could be a fantastic idea, and reasons that this could be terrible. Let’s start with the fantastic.
Why Robots Are Better Than Skype
Let’s review some shortcomings of Skype, FaceTime, and similar applications when they are used for communicating with younger children.
Screens feel distant: The benefit of robots over screen-based communication is that they are physically present. A six-month-old might ignore mommy or daddy on the screen, but find parent-robot delightful. A more companionable parent “avatar” might help distant parents feel closer to their kids. The robot doesn’t need to be particularly complex — Skype-on-a-stick is a great way to begin —because just the fact that it can move through the space will change the way kids interact with it.
The kids fight over the phone: I have 2- and 4-year-old sons, and they both want to hold the phone when I am traveling. If I had a telepresence robot, I could control the view myself, either moving the base—ideally an omnidirectional robot so I can rotate in place—or moving an actuated head. No more fighting.
A child decides to watch Netflix instead: I don’t know who came up with the idea of putting every application you can think of on a single device, but as a parent of two, I hate you! One of the major drawbacks of using a phone or iPad to Skype with your children is that these smart device super-users (a.k.a. 2-year-olds) might not be of the wait-to-eat-the-marshmallow variety, and realize they are just a button-and-two-clicks away from Clifford the Big Red Dog. Attaching your device to a projector generally helps, but a single purpose telepresence robot (hopefully with lots of fake buttons so they can’t figure out how to turn it off) could also be the solution.
Kids don’t like to sit still: If your child doesn’t have the gift of the gab today, sometimes adult-like turn-taking conversations become dull and they might not talk to you for very long. Contrast that with a robot that can play tag, or chase a toddler across the house…. Yay! Let’s play tag! There could be a whole new market at the holidays for books addressing games you can play with your kids through telepresence robots.
Design Guidelines for Robot Co-Parenting
Technology can be used and misused, so here are some guidelines and precautions for using co-parenting robots for positive impact.
Robots should not replace real parents: This might seem like a ridiculous newspaper headline from a dystopian future, but it is important to understand here that the purpose of social and socially-augmenting robots is to connect people, not replace human connection. Social networks don’t replace in-person friendships and direct conversation, and similarly, parents should not use telepresence robots as an excuse to stay late at work every night or just watch sports/do their nails when they come home. The point is to make the times when you need to be away from your children more satisfying. I tell my kids that one of the hardest skills to learn in the world is sharing, and so it is with parents who split the time with their kids.
Think about size: The telepresence robots out there today are not optimized for children. Having adjustable heights down to 2 feet or so would be great. Maybe someone can partner with Fisher-Price to make it childproof — and hide the power/volume buttons. (Willow Garage, a fabled robotics company, now foregone, had a remote worker who would drive into people’s chairs when people turned his volume off. I’m sure a child would find such a sequence entertaining too.)
Don’t use robots for alienation of affection: Maybe I’m biased, but I think robots are pretty cool, and a lot of kids do too. When someone goes out of town, that’s often a chance for the other parent to be in charge — so maybe don’t interrupt homemade pizza night with a super cool robot just when the kids were about to put on the toppings. I’m not sure exactly how to incorporate this into a robot user interface or behavior system, but it would be a fascinating area for future research. How do we get robots to help us help ourselves? Sometimes we know the right thing to do, but a little notification here or there could make us pause for a second before we hit that robot call button.
Don’t hack your ex’s robot: I’m not a psychologist, but I am pretty good with technology. That being said, it seems pretty intuitive that hacking your ex’s telepresence robot is not going to benefit your co-parenting relationship. I suggest robot designers consider cyber-security a top priority in this sometimes high-emotion application context.
The Distant Future of Robot Co-Parenting
Let’s let our imaginations run wild and think about future risks and benefits should such a technology truly take off. I’m having fun here, but sometimes science fiction is exactly what inspires future technology designers.
Robots as mediators: Thinking back to the alienation of affection by robot (lawyers should have a field day with that one in custody battles), could a robot have a role in improving parent-to-parent or parent-to-child communication? This might not be as far-fetched as it sounds. Cornell researcher Guy Hoffman and colleagues studied mediated telepresence between conflict-ridden couples [pdf], including a robotic telepresence that cowered when someone spoke in angry tones. The robot’s reaction, at least in a user study context, did cue the participants to speak in kinder tones. In addition to reminding you when might be a bad time to call, they might be able to give you live feedback about a current call.
Applications for marriage therapy: Related to the above, perhaps annotated interaction logs (or extracted quotes?) could be sampled in marriage therapy or individual counseling to help people learn to improve their communication patterns. Obviously privacy would be an enormous concern, so perhaps the parents would need to approve the clips ahead of time, host them on their own devices, and they would be automatically deleted after the session. Proceed with caution! However, these robots would be transmitting lots of personalized data, and just like with data-driven analysis, which is helping us solve problems from image classification to financial markets, I would expect future therapy practices to benefit from life-habits analysis.
What about privacy? With any technology that stores personal data, and as we have seen with the various hacked toys so far, co-parenting robots would need to have bulletproof data-protection. Just like when people thought they would never use credit cards on the internet, and now 75 percent of us do our holiday shopping online, people will get used to managing data privacy for social devices and general technology. We’re too addicted to them to not make them work. Robots with social capabilities also raise particular privacy concerns: If the robot had some local intelligence or conversational capabilities, it might easily convince a small child or the other parent to reveal personal information as they bond with the system itself. (So much more to say on this topic, but let’s save robot social manipulation for a future post.)
Remote co-parenting could leverage embodiment to help remote parents feel closer to their children via spatial games, singular use devices, and improved engagement. Its risks include bad physical designs (in terms of both safety and interaction), misuse by parents who meant better but find it easier to call via robot than to come home early, or other parents who also use it to stalk their partner (“Whose pair of shoes are those behind that door?”). The future is exciting, and with balanced consideration, I hope robot designers consider all sides.
And yes I am in Silicon Valley for the summer, so if there’s a company that wants to buy this idea and pay me lots of money to help them get it to work, I’m happy to take this post down.
But right now, it’s my children’s bedtime…
Dr. Heather Knight is currently the robotic artist in resident at X, the advanced technology lab of Google’s parent company, Alphabet. Starting in fall 2017 she will be a computer science professor at Oregon State University, in Corvallis. Her research interests include human-robot interaction, non-verbal machine communications, and non-anthropomorphic social robots. She completed her PhD in robotics at Carnegie Mellon University, and holds B.S. and M.S. degrees from MIT. Follow her on Twitter: @heatherknight