Can Robots Form ‘Relationships’? The Birth of Socially Assistive Robotics
When most people think of robots, they picture machines on factory assembly lines or devices performing dangerous tasks. Yet at the University of Southern California (USC), a fundamentally different kind of robotics research has been underway for decades. It is called Socially Assistive Robotics (SAR)—a field dedicated to driving cognitive, emotional, and behavioral change through social interaction between humans and robots. In April 2026, IEEE Spectrum published an in-depth feature spotlighting the USC pioneer who founded this field, examining the present and future of SAR research.
What Is Socially Assistive Robotics, as Pioneered by the USC Professor?
Socially Assistive Robotics is a research discipline in which robots help people through social and emotional interaction alone—without any physical contact. Robots serve as coaches, companions, and therapeutic aides for a wide range of vulnerable populations, including children with Autism Spectrum Disorder (ASD), patients with Alzheimer’s disease, stroke rehabilitation patients, and the elderly.
The robots developed by the USC research team combine a variety of multimodal sensing technologies—including facial expression recognition, voice analysis, and gaze tracking—to assess a user’s emotional state in real time and deliver appropriately tailored responses. The core goal is not simply executing commands, but forming a genuine ‘relationship.’
“The goal of a socially assistive robot is not physical assistance. It is for the robot to motivate people, change their behavior, and improve their quality of life through interaction.” — IEEE Spectrum, interview with USC professor
Key Research Achievements: From Children with Autism to Elderly Care
Therapeutic Support for Children with Autism Spectrum Disorder
One of the most notable achievements in SAR research is its application in therapeutic support for children with autism. When children find it difficult to make direct eye contact with other people, the robots developed by the USC team help them practice social skills first through interaction with the robot itself. Multiple studies have been published showing that robots provide children with autism a sense of psychological security, as their responses are more predictable and consistent than those of human therapists.
Cognitive Rehabilitation and Emotional Support for the Elderly
Both the United States and South Korea are aging societies, and demand for elderly care robots is growing explosively. USC’s SAR robots engage Alzheimer’s patients in memory training games and hold everyday conversations with isolated seniors living alone, serving as emotional companions. Researchers reported observing a tendency toward a slower rate of cognitive decline among groups of elderly individuals who regularly interacted with SAR robots.
Stroke Rehabilitation Coaching
The rehabilitation process following a stroke is repetitive and tedious, leading many patients to abandon treatment midway. SAR robots contribute to improved rehabilitation adherence by providing continuous encouragement and feedback at the patient’s side. In situations where a human therapist cannot be present around the clock, the robot fills that gap.
Technical Challenges: The Convergence of AI and Robotics
As SAR research grows more sophisticated, the level of technology required continues to rise. In the latest research, the integration of large language models (LLMs) with robot platforms has emerged as a central challenge. Robots equipped with GPT-class models are now capable of sustaining more natural conversations, but this has simultaneously introduced new challenges, including hallucination problems and data privacy concerns.
Because the USC team’s robots serve vulnerable populations, reliability and safety are treated as the highest priorities. If a robot conveys incorrect information or responds in an emotionally inappropriate way, it could cause serious harm to the patient.
Implications for Korean Readers
South Korea is aging at one of the fastest rates in the world, while simultaneously facing a severe shortage of caregiving personnel due to its declining birth rate. Socially assistive robots are drawing attention as a promising technology capable of addressing both problems at once.
Domestic research on emotion-aware robots is being conducted in Korea through institutions such as KIST (Korea Institute of Science and Technology) and ETRI (Electronics and Telecommunications Research Institute), but there remains a significant gap compared to USC in terms of clinical application and commercialization. In particular, the field of therapeutic support robots for children with autism faces immediate demand in special education settings across the country, yet R&D investment and clinical trial infrastructure remain insufficient.
Furthermore, in Korea, discussions around national health insurance coverage and government subsidy policies for elderly care robots are still in their early stages. Experts have called for more proactive policy design informed by USC’s research outcomes.
Conclusion and Outlook
The field of Socially Assistive Robotics, pioneered over decades by a USC professor, has now grown beyond purely academic research into practical technology being deployed in real hospitals, nursing facilities, and special education schools. Advances in AI—particularly the integration of LLMs—are opening new possibilities for SAR robots to communicate with people in ever more natural and effective ways.
Of course, ethical issues—including excessive emotional dependence on robots, privacy violations, and the allocation of medical liability—must also be resolved. When technology and ethics are held in balance, socially assistive robots will be able to transcend the limitations of human caregiving and take their place as true ‘social companions.’ South Korea, too, must closely monitor this trend and move swiftly to bolster research investment and build the necessary institutional foundations.
📚 References (1)
※ This article was written by synthesizing and analyzing the sources listed above.
Generated: 2026-04-22 18:01
AI & Robotics Newsletter
Subscribe for English AI & Robotics news every Mon & Thu.