An international team is pioneering the future of human-robot interactions, focusing on enhancing the way robotic companions interact with their owners through a combination of AI and edge computing, known as edge intelligence.
This project, backed by a one-year seed grant from the Institute for Future Technologies (IFT)—a collaboration between the New Jersey Institute of Technology (NJIT) and Ben-Gurion University of the Negev (BGU)—aims to transform robotic companionship.
Assistant Professor Kasthuri Jayarajah from NJIT’s Ying Wu College of Computing is leading research to develop a socially assistive model for the Unitree Go2 robotic dog. This model will enable the robot to adapt its behavior based on the characteristics of the individuals it interacts with, providing a more personalized and dynamic interaction.
The primary goal is to make the robotic dog seem “alive” by integrating wearable sensing devices capable of detecting physiological and emotional stimuli linked to a person’s personality traits and transient states, such as introversion or levels of pain and comfort.
This innovation has significant potential for home and healthcare settings, particularly in combating loneliness among the elderly and aiding in therapy and rehabilitation.
Jayarajah’s initial research, where robotic dogs interpret and respond to gestural cues, will be showcased at the International Conference on Intelligent Robots and Systems (IROS) later this year.
Co-principal investigator Shelly Levy-Tzedek, an associate professor in the Department of Physical Therapy at BGU, brings her expertise in rehabilitation robotics to the project, particularly her research on how age and disease affect body control.
The researchers highlight the growing accessibility of wearable devices, which can now extract data on brain activity and micro expressions using everyday items like earphones. The project aims to integrate these multimodal wearable sensors with traditional robot sensors, such as visual and audio, to track user attributes objectively and passively.
Jayarajah notes that while socially assistive robots hold great promise, their long-term use is challenging due to costs and scalability issues. “Robots like the Unitree Go2 are not yet equipped for large AI tasks. They have limited processing power, memory, and battery life,” she explained.
The project’s initial phase involves enhancing traditional sensor fusion techniques and exploring advanced deep-learning architectures to develop wearable sensors that can accurately extract user attributes and adjust the robot’s motion commands accordingly.
Related Article: New Land Predator Found Among Sea Creatures
