Effects of Interfaces on Human-Robot Trust: Specifying and Visualizing Physical Zones

Robots are slowly becoming a part of our lives and are interacting with us nearly daily. Trust is a necessity for humans to feel comfortable around robots. The mode of spatial communication between these machines and humans also affects underlying trust between people and robots. 

Atlas robot from Boston Dynamics. Image credit: Boston Dynamics

Marisa Hudspeth, Sogol Balali, Cindy Grimm and Ross T. Sowell have discussed this aspect in their research paper titled “Effects of Interfaces on Human-Robot Trust: Specifying and Visualizing Physical Zones” which forms the basis of the following text.

Importance of Human-Robot Trust

Robots can make our life easier and efficient. Hypothetically, Robots can efficiently complete all manual tasks for us. For the smoother integration of robots in our lives, a high trust relationship is necessary.

Often, trust between humans and robots is affected by the mode of communication between them, and this research analyses how different interfaces affect the underlying trust in human-robot interaction.

Research Methodology

Robot Human interaction can be segregated into two sub-groups:

• Interface: Human communicating to the robot
• Feedback: Robot communicating to the Human

In this specific case, the researchers have considered, a human communicates to a Robot that a particular area in a room is a no-go zone which means that a robot should not enter this space. This communication can happen in the following three ways:

• Physical Interface: Placing cones across the room in such a way that it forms a quadrilateral which is a no-go zone for the robot
• Map Interface: Marking an area in a map as a no-go zone
• AR Interface: Using an Augmented Reality interface to mark an area as a no-go zone

The robot communicates the no-go zone to the human by:

• Physical Feedback: Avoiding the no-go zone
• Map Feedback: The area is marked as RED on the map
• AR Feedback: A virtual fence is displayed on the smartphone’s screen
• No Feedback: To evaluate the importance of Feedback on trust.

Types of various interfaces and feedbacks discussed in this research work. Image credit: arXiv:2112.00779 [cs.RO]

How is the Trust Evaluated

Trust can be assessed by the below three means:

• Method1: We ask the human participants to fill a form that asks them how much do they trust the robot not to enter the no-go zone
• Method2: It could be indirectly derived based on the willingness of participants to put breakable objects in the no-go zone.
• Method3: 2 tables can be placed inside the no-go region at varying distances, near and far. The participants were asked to put durable/fragile items on these tables such that the robot would not knock them down.

Research Result

Participants preferred a different modality for feedback than the interface:

• Participants preferred to use the augmented reality interface to specify a no-go region.
• Participants preferred to receive physical feedback.

The research paper can be referred to for detailed experiment methodology and the results of the experiment

Conclusion

The researchers concluded that interface and feedback play an essential role in the trust level in Human-Robot Interaction (HRI). We can use the results obtained from this research to design effective interface and feedback mechanisms for Human-Robot Interaction. This understanding will help us facilitate an environment of trust for humans while interacting with robots in their physical space. In the words of the researchers,

In this paper we investigate the influence interfaces and Feedback have on human-robot trust levels when operating in a shared physical space. The task we use is specifying a “nogo” region for a robot in an indoor environment. We evaluate three styles of interface (physical, AR, and map-based) and four feedback mechanisms (no feedback, robot drives around the space, an AR “fence”, and the region marked on the map). Our evaluation looks at both usability and trust. Specifically, if the participant trusts that the robot “knows” where the nogo region is and their confidence in the robot’s ability to avoid that region. We use both self-reported and indirect measures of trust and usability. Our key findings are: 1) interfaces and Feedback do influence levels of trust; 2) the participants largely preferred a mixed interface-feedback pair, where the modality for the interface differed from the Feedback.

Source