I recently watched an introduction to Google AIY by James McLurkin, Google head of hardware design, and Limor Fried, owner of Adafruit Industries. They discussed ways of using new AI-embedded devices and their motivations for the project. This has direct relevance to household and social robot design.
The speakers said their first motivation is letting people train Google AI recognition in their homes. The second, according to McLurkin, is “the top-secret plan is to build things that build more engineers.”
Spoiler alert: Engineering won. The very complex ecosystem we are developing is getting incredibly difficult to understand without been trained in science and mathematics. The left-brained part of humanity is responsible for the most part of our progress. With the help of data, our understanding of the universe is growing, and nothing seems impossible anymore. However, a purely technical approach to social robot design is unlikely to succeed.
Perception versus reality
Robotics represents a great example. The field has experienced steady progress, thanks to accessible technologies and lower prices. Boston Dynamics’ SpotMini has received strong interest — and fear — from the public via social media.
Oddly enough, the public conception of robots doesn’t fully appreciate the technical advances in products designed to perform specific tasks. I hope that the rise of collaborative robots will give designers more opportunities to create cobots that work side by side with humans.
But what about social robots? They are not as precise as their industrial cousins. What is the main purpose of a companion robot? Developers and vendors claim it ranges from being a personal assistant to being a nanny, a coach, a teacher, an aide for children with disabilities, or anĀ elder care worker.
But what is the main function of robotic companionship? In that case, the rational measures of industrial automation effectiveness are not adequate for subjectivity of the service market.
SoftBank Robotics’ Pepper has been used in numerous public locations such as hospitals, airports, and retail stores, but contract renewals for the humanoid robot have reportedly been weak.
The audience is a barrier where reality strikes back against limited robot design. People are not attached to a function but to a set of benefits. Consumers want a brand with a meaningful message that both challenges their expectations and enables them to project their own preferences. The market is the ultimate judge of success.
To quote Zander Nethercutt, “People don’t buy products; they buy better version of themselves.” This is even more important for social robots, which can be costly and could be considered as luxury products. Unfortunately, there are no engineering formulas to solve that problem, and maybe the industry could try an alternative road.
Time for a new approach to robot design
It is time to explore of a complementary approach. Engineers and designers should work together to creatively explore a concept holistically. The strength of engineering is having team members solve narrow problems. On the other hand, designers can draw a subjective map of a robot design, allowing multiples options and scouting for hidden angles.
The designer’s role consists of more than improving the cosmetic appearance of a product. Good design must also include how a robot integrates into a home (through interior design and smart home technology) and what customers need (via inclusive design).
An example of successful robot design is the PARO therapeutic robot developed by AIST in Japan. The interactive device is designed to look like a seal and has helped relieve stress in dementia patients.
Another example is KUKA, whose longtime collaboration with industrial designer Mario Selic has paid off. The company has won the prestigious Red Dot award several times in different categories.
Household and social robots will be very complex creatures. Approaching the domestic market as a problem/solution paradigm is doomed to fail. Redundancy is a product killer, as seen in the recent demise of Jibo. Ubiquitous tech already offers instantaneous access to information, so who needs to open e-mail on a remote screen when your phone is constantly within range?
In addition, Jibo was fundamentally flawed in not being designed around users. Many people didn’t know what to do with it in their homes.
While general-purpose robots are still a ways off, consumer models need to be more than a tablet on wheels or a swivel stand. Solving this robot design challenge will require designers and engineers to work closely together on subjective and objective goals.
The Massachusetts Institute of Technology is a leader in industrial design. Looking back a few years, MIT’s City Home project used robotics as a means to an end, that is, seamlessly facilitating life in a domestic environment.
The principles of good design start with a meta-analysis of what the user needs, how, when, and where. Hardware, software, and the user interface are all part of that, and it is a delicate ballet in need of a choreographer.
Teams working on social and domestic robots should use state-of-the-art technology, build their brands carefully, and use agile methodology to communicate. That way, they can reconcile the right and left sides of our brains and build better robots.
About the Author:
Alexandre Colle is an industrial designer and a design consultant specializing in social robotics. He recently graduated from Central Saint Martins in London with an MA in industrial design.