Let’s talk robots. They may not be too noticeable, but they are all around. Your virtual, voice-activated assistants, like Siri and Alexa, are constantly answering queries, both mundane and serious. Look down and you may see a Roomba robotic vacuum cleaner mapping your floor while cleaning it.
Covid-19 has brought humans even closer to robots. At the height of the pandemic, various kinds of robotic machines and systems have been deployed everywhere, from supermarkets to hospital corridors, seamlessly filling the gap between people that opened up due to the new coronavirus.
Expect more such “working” robots in the years to come. In a new book, What To Expect When You’re Expecting Robots (Hachette, 304 pages, Rs1,525), roboticist Julie Shah and co-author Laura Major explore how the robots of the future will not work for us but with us—in tandem. Shah is an associate professor of aeronautics and astronautics at the Massachusetts Institute of Technology, while Major is the chief technology officer of Motional, a self-driving car venture supported by automotive companies Hyundai and Aptiv.
In the book, they envision a future where robots and humans not only work as individuals, “buzzing around each other”, but also collaborate in groups. “People say when you are trying to work on human-robot interaction, it’s almost like a cheat—you are giving up on the challenges of robotics and trying to use the person to fill in those gaps,” says Shah in a video call. “Neither of us came with that view. Instead, we said what are we developing this technology for if it’s not to enhance our capability and well-being?”
Quite simply, human-robot collaboration will be a “two-way street,” write Shah and Major. They say this partnership will require new human and robot languages. Robots will need to be designed so that they understand social norms. Humans, in turn, will have to rethink the role of technology, and rethink infrastructure to incorporate the new intelligent robots. “I think the way we are going to interact with them needs to be humanised,” says Shah.
The authors illustrate this by comparing a robot to a student driver. Experienced drivers instinctively know how to change their driving behaviour when they see the “student driver” or “learning” sign on a vehicle. There are rules and modes of behaviour that only come with experience. Student drivers do not have these mental models.
Shah and Major ask readers to imagine driving every day while surrounded by student drivers. “Now imagine how exhausting and stressful—and possibly dangerous—it’s going to be to coexist with hundreds of robots in our lives every day, whether on roads, in the hallways of our office buildings, in our parking lots, in our restaurants or hospitals…especially if they don’t understand the rules we all follow that make those spaces navigable and safe,” they write.
Robots, then, will have to be designed differently. The usual rules for consumer electronic devices won’t work. Anthropomorphising robots or making them more relatable won’t be enough. “Robots don’t have to be cute,” notes the title of a chapter.
Moreover, adds Shah, system safety nets that allow humans to intervene when needed will be required.
Take autonomous vehicles. Today, you can expect to get into a pod and just be driven to your destination. “To the extent that we accept that these technologies are not going to be flawless or infallible.... We can no longer aim for the goal where we never have to take over,” Shah adds. “We know that from aviation. If a pilot remains out of the loop for long enough, you undermine their ability to perform the critical role of stepping in when the unanticipated happens and take control of the aircraft,” Shah says.
Be it delivery robots, security guard robots, or robots for critical tasks in offices, hospitals and other workplaces, Shah believes they won’t necessarily have to look humanoid. But they would have to strike a fine balance between safety, productivity and aesthetics.
The book also explores the possibility of robots talking with one another. Today, as the authors write, we can barely coordinate two autonomous cars at a standard intersection. In the future, a fleet of delivery drones and robots will be able to communicate with each other to avoid collisions or ensure there aren’t too many of them in one block. Where do humans fit? They will merely be the conductors or supervisors stationed at command centres to ensure the fleet of robots works in harmony.
Till then, the Roombas in our homes and the supermarket robots mopping up spills are performing their tasks just as they are meant to. “They are doing a very narrowly defined task and they do it well. But we do have the capability to make these systems more capable in less-structured or less-controlled environments,” says Shah.