A social robot, defined as a robot designed for human interaction and to evoke emotional connections, strives to make interactions feel akin to engaging with "someone" rather than a mere "something." An ideal social robot is perceived as having autonomy over its actions, making seemingly independent choices, and acting autonomously. These robots achieve this "someone" quality through artificial intelligence (AI), using data from cameras and sensors to make informed decisions.
Image credit: Have a nice day Photo/Shutterstock
Recent AI advancements enable the translation of psychological and neuroscientific insights into algorithms that empower robots to recognize voices, faces, and emotions, interpret speech and gestures, maintain eye contact, converse naturally, and adapt based on feedback.
Making Social Robots
AI faces the core challenge of creating machines capable of seamlessly and successfully interacting with humans. This problem is often viewed as the need for a highly adaptable interface between artificial and human cognition. Human intelligence is inherently collective, leading to the development of language, conventions, norms, belief systems, technologies, and organizations that surpass individual capabilities. Human cultures and civilizations exhibit remarkable diversity and adaptability. Therefore, the true goal of AI should be to create machines capable of contributing to collective human intelligence by participating in conversations and projects.
Overcoming substantial obstacles is essential to achieve a truly 'social robot' or an AI system capable of effective collaboration with humans. While there have been impressive advancements in creating life-like interactions between AI and humans, the deeper challenges of collaborative intelligence have not been adequately addressed. Human communication involves a level of flexibility that goes beyond language, including pragmatic inference and joint reasoning.
To create a truly social robot, one must consider processes such as "virtual bargaining," where AI agents imagine agreements that could result from hypothetical negotiations with other agents. This not only influences their behavior but also anticipates the actions of others. Such virtual bargaining leads to the development of complex systems of rules, conventions, norms, ethics, knowledge, and institutions akin to the principles of spontaneous order observed in both the natural and social sciences.
However, implementing a virtual bargaining robot presents conceptual and computational challenges, demanding a deeper theory of bargaining and an understanding of how hypothetical bargains can be simulated to align multiple agents' interests.
Applications of Social Robots
The core of social robots lies in their algorithms and AI systems. Machine learning and deep learning techniques are applied to recognize human activities, emotions, and behavior. These AI systems enable robots to respond empathetically and engage users effectively. These robots go beyond simple tasks, taking on roles in customer service, caregiving, and companionship.
The potential for robots to excel in caregiving roles has led to research targeting both ends of the age spectrum, aiding children in managing anxiety and enhancing social skills. They also support the elderly with medication reminders, physical assistance, and companionship.
Social Robots in Healthcare: Continuous technological advancements that expand their utility and impact drive the increasingly crucial role of social robots in healthcare. These robots offer emotional support, which can significantly impact patients' well-being. Beyond emotional support, they enhance healthcare by improving accessibility, efficiency, and cost-effectiveness. Their ability to alleviate anxiety, depression, and other mental health conditions has been well documented. Social robots have been employed successfully in various medical domains, including psychiatry, pediatrics, geriatrics, and rehabilitation.
Social Robots as Educators: Robots have found a prominent role in education, primarily serving as tools to teach STEM subjects. Initially used to introduce programming and computational thinking, they have expanded their utility to include electronics, mechanical design, the arts, and the development of soft skills such as collaboration and negotiation. This evolution has been studied for over five decades, showcasing its effectiveness in various educational settings.
Social robots hold promise due to their lifelike behavior and ability to engage learners effectively. They offer a cost-effective means to complement teachers by tutoring small groups or individual learners. Research suggests that tutoring, whether by humans or robots, is highly effective in education.
Social robots excel in social interaction, providing the physical and social presence missing in traditional computer-based learning. Their tangible nature enhances engagement, compliance, and motivation. While interactions with robots differ from human interactions, they evoke social responses and establish connections.
They serve as tutors, offering support to individual learners or small groups, and provide personalized attention, assisting students falling behind or challenging those ahead. With infinite patience, robots can repeat lessons until mastery. They also create a non-judgmental learning environment, reducing learner anxiety.
Ethical, Privacy, and Trust Issues
Ethical Considerations: In the healthcare domain, social robots play a significant role in assisting vulnerable individuals, such as the disabled, children, and the elderly. These robots, equipped with AI, exhibit a high degree of interactivity. However, this very aspect introduces ethical challenges. The central issue revolves around the potential dehumanization of interactions when vulnerable individuals are cared for by entities devoid of emotions or an understanding of human fragility. The replacement of human caregivers with robots raises ethical concerns.
To address these concerns, developers have implemented algorithms capable of recognizing emotions and facial expressions from sensory data. This improvement enhances robot acceptance and usability, aligning to create socially adept systems. Additionally, public trust plays a pivotal role in accepting and deploying social robots. Trust increases when these artificial systems provide significant safety and well-being benefits.
Ethical considerations have become a prerequisite in the design of social robots. Stakeholder research at the outset of the design process is essential to understand when human-robot interaction is preferred, given robots' predictable behavior or their potential to optimize the time of nurses and physicians for the benefit of patients. The primary objective is to translate social norms, laws, and human behaviors into algorithms for intelligent artificial intelligence. Instead of replacing human caregivers, robots can act as companions to alleviate loneliness and mitigate dehumanization.
Privacy and Security: Social robots operate in both public and private settings, especially in healthcare, where they interact closely with people, including the elderly, disabled, children, and medical professionals. The involvement of robots in various scenarios with diverse end users necessitates data protection measures to safeguard individuals' privacy. Legal frameworks have imposed requirements and constraints on the secure management of data recorded by social robots, which often rely on wireless connectivity and cloud-based systems to store information.
Building systems that securely handle personal data requires a foundation of 'privacy by design' in social robot development. Blockchain technology holds promise for ensuring both privacy and security. However, this innovative solution requires further research and validation for practical application. Currently, blockchain technology faces a trilemma involving security, performance, and decentralization, which cannot all be simultaneously achieved.
Trusting Social Robots: Evidence suggests that social robots play an increasingly prominent role in various aspects of life, such as healthcare, entertainment, and personal support. People tend to develop relationships based on trust with these robots. An example is the therapeutic healthcare robot PARO, designed to assist elderly individuals, particularly those with dementia. During the COVID-19 lockdown, studies indicated that lonely individuals found comfort and companionship in robot interactions. Interactions with robotic dogs significantly alleviated loneliness, with minimal differentiation from interactions with live dogs.
Trust can develop based on how social robots present themselves, even when their behavior is intentionally designed to create false beliefs in humans. While this appearance-based trust facilitates relationships, it also questions the appropriateness of trusting objects that present themselves deceptively. Trust may be directed toward a facade, potentially challenging its foundation.
References and Further Readings
Breazeal, C., Dautenhahn, K., and Kanda, T. (2016). Social Robotics. In: Siciliano, B., Khatib, O. (eds) Springer Handbook of Robotics. Springer Handbooks. Springer. DOI: https://doi.org/10.1007/978-3-319-32552-1_72
Ragno, Luca, Alberto Borboni, Federica Vannetti, Cinzia Amici, and Nicoletta Cusano. (2023). Application of Social Robots in Healthcare: Review on Characteristics, Requirements, Technical Solutions. Sensors 23, no. 15: 6820. https://doi.org/10.3390/s23156820
Cross E. S., Hortensius R., and Wykowska A. (2019). From social brains to social robots: applying neurocognitive insights to human–robot interaction. Philosophical Transactions of the Royal Society. B 374:20180024. DOI: http://doi.org/10.1098/rstb.2018.0024