An innovative humanoid robot, iCub 3, is breaking new ground by enabling users to experience events remotely using haptic feedback gloves and a virtual reality (VR) headset. Weighing 52 kilograms and standing at 125 centimeters tall, the iCub 3 boasts 54 points of articulation, with cameras and sensors strategically placed across its aluminum alloy and plastic body. These sensors send data to the robot’s “brain,” allowing it to replicate sensations on a suit and VR headset worn by a person located hundreds of kilometers away.

Equipped with two cameras in its head resembling human eyes and an internet-connected computer functioning as its brain, the iCub 3 captures visual data and transmits it to the remote operator. The suit worn by the operator also features sensors that pick up their movements, enabling the robot to mirror their reactions. While there may be a slight delay of up to 100 milliseconds in capturing and transmitting visual footage, users can adjust by moving slightly slower than normal.

The iCub 3 has been showcased at the Venice Biennale, where it navigated through an exhibition while its operator stood 290 kilometers away in Genoa. Stefano Dafarra, a member of the iCub 3 team at the Italian Institute of Technology, emphasizes the importance of translating every signal and numeric data sent through the network for seamless communication.

Dafarra envisions the iCub 3 as a tool for remote event attendance, minimizing the need for physical travel. However, concerns persist regarding the robot’s vulnerability to falls and its ability to stand up independently after such incidents. Despite these challenges, the iCub 3 represents a significant advancement, offering advantages over its predecessors.

While acknowledging the robot’s potential, Jonathan Aitken at the University of Sheffield expresses disappointment in the lack of clarity regarding the data transmission requirements of the new version. He emphasizes the importance of understanding the amount of data required and establishing upper and lower bounds for efficient operation.

By Impact Lab