Communicating with Social Robots | KPMG | DK

Communicating with Social Robots

Communicating with Social Robots

For a realistic and durable human-robot interaction, social robots need human-friendly appearances, communications and cognitive skills.



Senior Consultant

KPMG in Denmark


Related content

Social robots need human-friendly appearances, communications and cognitive skills

Co-written by: Evgenios Vlachos, PhD, Aalborg University & Simren S. Dhaliwal, Senior Consultant, KPMG NewTech

Communication between humans can be difficult enough to master, but making a social robot join the conversation brings a whole new range of challenges to the table. Social robots are able to interact and communicate with humans in a human-like manner, all while respecting the existing social and cultural norms of its environment.

Take Pepper, the social human-shaped robot from SoftBank Robotics.  A typical example of a robot, designed to be a genuine day-to-day companion. You can compose music with it, let it play with your kids or use it as your wedding planner. Pepper's number one quality is the ability to perceive emotions and adapt its behavior to the mood of its human counterpart.

Applying social rules
When interacting with such robots we as humans apply social rules, and act on inherited behavioral guidelines, expecting that the robots will have the ability to understand, and follow them. We have adjusted our space, actions and performed tasks to what we are physically capable and incapable of.

In the same way, the features of a social robot should fit within these predetermined boundaries when it is beneficial for the user or for completing a task. The robot should not fall over when it is practicing dance moves with you or talk with a loud angry voice when reading a lullaby for your kids in the evening.

A human-friendly appearance 
For a realistic and durable human-robot interaction (HRI), social robots should incorporate human-friendly appearances, communication functions, user-behavior modeling techniques and cognitive skills, as well as the common characteristics of industrial robots, namely: Accuracy, carrying capacity, durability, repeatability, safety and speed.

The variety of social robots and the number of interfaces for HRI have grown rapidly in recent years. More than ever, we have to investigate which robot types can best satisfy the requirements of a given task. Tasks for social robots are usually related to healthcare, wellness, assistance, entertainment, companionship, and education. Some examples include the android Geminoid-DK taking up the role of a university lecturer, the teleoperated humanoid (Telenoid) facilitating communication with elderly people suffering from dementia and the humanoid robot (Kaspar) promoting cooperative play among children with autism. Social robots may also have a zoomorphic (animal-like) appearance, like Sony’s robot dog AIBO, or even a mechanoid (machine-like form like Cynthia Breazeal’s Jibo, depending on what suites the task it was created to solve.

Apart from very few occasions where one can encounter a social robot “in the wild”, such as the use of Pepper at Mizuho Bank in Japan, most social robots are still confined to laboratories, research facilities, nursing homes and museums.

Key considerations when communicating with robots
Until social robots become a naturally integrated part of our society, here are some key considerations to keep in mind when communicating with such robots – equally critical for those considering buying or building them:

  • Failure to meet the expectations that rise from the appearance, actions, perceived intelligence, or perceived gender of a social robot might lead to disruption of the interaction - the robot needs to be as capable as it appears to be.
  • Robot appearance matters less if the situation is engaging. With an engaging scenario, users react emotionally to the robot as if it was a human, and do not lose focus of attention by possible faults in the appearance, possible lack of expressions, or technical drawbacks.
  • Social conditioning can render people reluctant to interact with a robot of very human-like appearance.  For example, a very human-like robot looking continuously at the speaker may not be a very natural behavior as mutual gaze happens somewhere between 30-60% of the time in interpersonal communication. Looking continuously may be perceived as starring by some people.
  • The physical location of the robot significantly affects people's expectations for the robot interaction. People expect different things from robots located in, for example, a manufacturing plant versus a storefront. The robot solution should reflect these differences.
  • Interaction with more than one social robot can affect the interaction experience for each of the robots. When diverse types of social robots (with different appearances and properties) are operating within the same area, the preferences and attitudes of users against a specific robot can be significantly influenced by interactions with the other robots. Similarly, bystanders who observe others interact with a robot may be affected towards interacting with the robot.
  • Touch (haptic communication) may be used as a way of inducing trust towards the robot.

© 2017 KPMG P/S, a Danish limited liability partnership and a member firm of the KPMG network of independent member firms affiliated with KPMG International Cooperative (“KPMG International”), a Swiss entity. All rights reserved.

Connect with us


Request for proposal