Predictability key to robot-human collaboration, finds Disney research
human-robot intereactions - disney research

Predictability key to robot-human collaboration, finds Disney research

Disney has published new research on human-robot interaction.

The research’s focus is to explore how to make robots move in a way that encourages human beings to cooperate with them, and to develop better human-machine rapport.

Part of human nature is to make snap judgements based on aesthetics. It’s a trait with obvious evolutionary benefits: determining threats before they get too close aids any species’ survival. This is why robot posture and movement will be increasingly significant for their future development and acceptance in the workplace.

In short, nobody wants to work alongside, or interact with, machines that look intimidating or dangerous. This is why some humanoid robots, such as Aldebaran/SoftBank’s NAO and Pepper machines, are designed to be small, non-threatening, and almost childlike, and to have appealing personalities.

The manufacturers’ aim in those cases is to make human beings feel safe around the robots, and even to feel protective towards them: a counter-intuitive idea that’s designed to counter the ‘Terminator problem’ of human beings feeling threatened by lifelike machines.

A similar principle applies to all types of robots, including ‘cobot’ machines that are designed to work alongside people in factories or warehouses, and even in hospitals or restaurants. Which is where Disney’s research comes in.

Read more: Flipping hell! Flippy the burger-bot too fast for human workers

Reading robot body language

Disney’s research explores human-to-robot handovers by varying a robot’s body language. The team wanted to examine how changes in a robot’s behaviour, posture, and movement, would influence human attitudes to the machine, and desire or ability to work alongside it.

The study found that the robot’s initial pose made a difference to the fluency and efficiency of the overall handover of a task, either from machine to human, or vice versa. The speed of the robot and its grasping method also had an impact on how well human beings felt able to interact with it.

According to the researchers, “This effect may occur by changing the giver’s perception of object safety and hence their release timing. Alternatively, it may stem from unnatural or mismatched robot movements.”

The key to more fluent interactions? Making the robot predictable, according to Disney. Participants in the study also reported less discomfort and more “emotional warmth” as they became more familiar with the robot’s behaviour.

“We find these results exciting, as we believe a robot can become a trusted partner in collaborative tasks,” wrote the researchers, Matthew Pan and Elizabeth Croft of the University Of British Columbia and Monash University, respectively.

Read more: DHL US trials robots, AI, AR & crowdsourcing to beat Amazon

Building rapport with robots

The most obvious conclusion of the study is that it’s possible for humans to build a kind of rapport with their robotic counterparts. While it might seem superficial or even unlikely, it seems to be key to developing effective human/machine partnerships – particularly when carrying out manual tasks.

Research such as this could lay the foundations for best practices in robot design and manufacturing. Making robots ‘affable’ and predictable is a big step toward convincing human workers that lending an artificial hand is something to be embraced, not intimidated by.

Internet of Business says

The need for human beings to feel safe around robots, and even to enjoy working with them, will be a vital consideration in the future, as robots of every kind move into the workplace. If ‘cobots’ are to collaborate with people successfully in manual labour settings, of any type, then human beings must feel valued and understood, and equally, must value and appreciate their machine colleagues.

The same principle applies to humanoid robots, and perhaps especially to ones that are designed to both understand and express or simulate human emotions.

A seminar at London’s Design Museum in January 2018, jointly presented by psychologist and writer Adriana Hamacher, and Internet of Business editor Chris Middleton, shared a number of research findings about how people will form bonds with humanoid robots, and will even lie to an emotion-expressing robot in order not to hurt its ‘feelings’.

Hamacher’s own research programme was reported here in 2016, and proves that emotional bonds between humans and machines can be genuine, and often surprising. Her study concluded that people often prefer working with robots that are expressive to ones that are merely efficient.