What we do
Research and Innovation
TCS Research: Inventing to build sustainable futures
Highlights
The semi-autonomous prototype allows a robot to move by itself, with a goal point specified on the map.
Additional features like automatic responses to voice conversations, text chat, and cognitive abilities with artificial intelligence, will be incorporated in the future, moving the needle closer to it becoming a fully autonomous robot.
In this article
Upgrading a telepresence robot during a pandemic
Pace Port™ Tokyo, in a mere matter of weeks, developed a prototype for a telepresence robot with semi-autonomous driving and path finding functions, despite hurdles presented by the pandemic and related restrictions at various levels.
The recently created proof of concept (PoC) has been a step toward Pace Port Tokyo’s efforts to introduce a fully automated robotic guide to conduct tours at the facility and interact with customers and clients via video conversations, and eventually replace the manually controlled robot currently used. The development was completed in about eight weeks in collaboration with the Rapid Development Team at Tokyo and TCS’ Cognitive Robotics research team.
A telepresence robot enables a human to be virtually present in a remote location through a tablet, smartphone, or a computer screen. The operator can interact with humans with the help of video camera, speakers, and other accessories.
Achieving semi-autonomous functions
The robot was previously operated only manually, where a human remotely controlled its movements and carried out video conversations with customers through the robot. The latest prototype allows the robot to move by itself, with a goal point specified on the map, paving the way for both parties to focus on the conversation.
Some of TCS’ Cognitive Robotics’ research team’s assets that made it possible to develop the semi-autonomous telepresence robot were edge-centric communication protocol, autonomous navigation using simultaneous localization and mapping (SLAM)—a technology that simultaneously estimates own location and maps its surroundings—and human-robot interaction using speech for dialogue exchanges and speaker localizations.
Not short of obstacles
But the quick development of the PoC was not short of obstacles. Besides the pandemic, one of the two major technical challenges faced by the teams involved was incorporating the semi-autonomous capabilities into a sandbox environment, due to the lack of common development and production environments.
Another challenge was achieving a sync between a low latency communication protocol that transmits messages and the video conference feature, due to differences in hardware infrastructure of the research lab and Pace Port Tokyo, and internet network compatibility.
Pace Port Tokyo plans to enhance the robot by incorporating voice-based navigation and advanced obstacle detection functions into this framework. The features implemented by the Tokyo team are promising in terms of what robotics and artificial intelligence can achieve for an enterprise in its journey toward digital transformation.
Favorable prospects
The team in Tokyo also believes that adding sensors in the future could help the robots perform more functions completely on their own, such as detecting and avoiding physical obstacles- animate or inanimate, or an automatic response to voice conversations, text chat, and cognitive abilities using artificial intelligence.
A wide range of possible uses of the telepresence robots that are already in demand include accompanying customers at commercial facilities, guiding them to their seats or rooms, and guiding them to the counter or examination room at hospitals. Such functions are especially useful during emergencies like the COVID-19 pandemic in order to reduce human contact as much as possible.
TCS’ Cognitive Robots research team is aiming to build robots that identify unknown environments and understand their instructions through language and gesture, toward enhancing human-robot interaction. Pace Port Tokyo is keen on introducing cognitive capabilities in their telepresence robot that would be helpful in industries such as education and healthcare.
The aforementioned functions of futuristic, fully automated telepresence robots, when combined with IoT have the ability to offer a truly virtual experience. Additional sensors could help pick up surrounding behaviors and allow users to interact with and respond to the robots in real time. Compatibility with third-party video conferencing capabilities could enable collaboration with wider stakeholders.
