Hands-on experience with social robots

As we look toward Society 5.0, it has become clear that the human-robot interactive ecosystem, especially for healthcare, must be carefully evaluated and designed considering the need for robust social policies—to augment the vision and create safe co-existence. ‘Social’ robots—a crucial part of this ecosystem—are meant to live with humans and share their social, personal, and interaction spaces.

Students from the University of Tokyo and Keio University in Japan got to work with a social robot for a month at TCS Pace Port™ Tokyo. TCS Japan provided this opportunity to students keen on exploring and experimenting with innovative solutions. Following an open call, five candidates of different nationalities—Japanese, Chinese, and Mongolian—were chosen for a month-long internship program through a rigorous selection process.

Prototyping at TCS Pace Port Tokyo

Social robots can take on many roles, such as personal assistants at homes and offices, nurses in the healthcare industry, companions for the elderly and children, and helpers in the hospitality sector. These robots have a great potential to improve human lives by assisting with physical, social, and psychological care. 

For the internship, three sub-themes within the larger forward-looking theme of social robotics were identified for exploration. These were robotics technology, robotics interface, and robotics-human interface. TCS Pace Port Tokyo, which offers TCS Pace™ capabilities, is an ideal space for interns to create prototypes, experiment with robots, and collaborate with the larger TCS Pace ecosystem. The interns developed various functions to make Misty, a programmable social robot, more human-like.

The interns were assigned two mentors, one based in Japan and the other in India. The mentors assigned them problems in the focus areas of design considerations for social robots and their potential uses and challenges in psychosocial care. The mentors also gave them tasks related to designing robots that mirrored human emotions to establish effective interactions, developing nursing assistants, and achieving shared autonomy in social robotics. 

The interns worked with Misty to develop and test their solutions. Dr Balamurlidhar and his team from the robotics research lab in India interacted with the interns daily to tweak any issues and share ideas to help with problem-solving. In Tokyo, Kuniko Takato and Nandhakumar supported them in refining their work.

Holistic research approach by multidisciplinary interns

Designing a robot encompasses determining its modes of interaction, types of social behaviors, and cognitive and physical abilities. These aspects help build acceptable and effective robots that fulfill their objectives in the long run and avoid exaggerated designs that leave the target users underwhelmed. 

The multidisciplinary backgrounds of the interns helped to take a holistic approach toward the designs and functions of social robots. They were able to provide inputs from both technology and social sciences perspectives.

Internship outcomes

The interns worked on Misty expressing emotions through eye and facial movements, making the robot understand and empathize with human emotions. The interns used machine learning to introduce functions such as obstacle recognition and avoidance to reach a destination safely, and researched the taxonomy of social robots and policies for design. Their findings revealed the need for new policies and initiatives to support social robots, especially for psychosocial care.

The interns also got an insight into TCS’ innovation ecosystem. At the end of the internship, the students presented their work—each intern with unique findings—to TCS Japan’s executive management. Their work would be absorbed into TCS’ robotics research team and deployed at the TCS Pace Port post further refinement to enhance customer experience and interaction.

The outcomes from this internship could also be modified and extended to assistive robots. For instance, to provide shared control capabilities to a pilot and an assistive robot, where the pilot imparts high-level inputs corresponding to a task and the robot performs the intended action using its manipulation capability. Such a system also finds applications in rehabilitation, mobility assistance, and active medical care.