In this project, we have created a theoretical framework for the design and development of a “needy” socially assistive robot (SAR) for the elderly. Our hypothesis is that a needy SAR will be more readily accepted by those who require it and would result in overall more positive care outcomes when compared to a non-needy alternative. This is supported by existing works that indicate traits that can be considered needy are viewed favorably when correctly integrated. We provide guidance on the key design considerations that should be made in order to implement these traits successfully into a SAR. Using this newly created framework we will subsequently proceed to evaluate the effectiveness of integrating a SAR with these traits in a real-world scenario.
Neediness. For a robot to be considered “needy” it must emulate a dependence on the user, such as by requesting that the user simulate the process of assisting or taking care of it. These requests do not need to be limited to speech synthesis, as neediness can be expressed in many different forms, such as through gestures or text. All needy interactions should be positive as perceived enjoyment has an
effect on the intention of elderly users to interact with a SAR.
Restraint. Neediness can be viewed as a negative trait unless presented in a specific endearing way, or when dependence becomes excessive and unmanageable. Therefore, a robot must not request more
support than is viable for the user to provide. An elderly or frail user for example must not feel pressure to perform actions when it would be uncomfortable to do so. In such a situation the robot should either stop prompting or find a different method of interacting with the user. While the user should be encouraged to interact with the SAR, there should be no significant downsides when failing to care for the robot, as this could cause undue stress. One of the most prevalent barriers to pet ownership is the physical demands of providing care overwhelming older individuals. In some cases this extends to individuals neglecting their own personal health and well being in an attempt to provide this care. It is important that the future development of needy SAR’s does not lead to similar problems.
Honesty. While anthropomorphism is important in developing empathy for a SAR, it must not attempt to deceive the user into believing that is a living being. Explaining that the SAR is a tool rather than a pet is preferred, particularly for the elderly or disabled, as this can be confusing and demeaning. They must
also not be presented as a replacement for human contact. Instead, they should be integrated as a method to afford greater independence to the user. Character. Robots should have a defined personality as they are viewed more positively than those who stick to formal means of communication. Interaction through friendly and casual means, such as using informal language, can greatly influence the enjoyment that users receive from using SARs. Consistent communication between users and robots forms social bonds between the two, resulting in greater satisfaction. Mistakes with long term consequences can drastically reduce trust [28], however expressing regret and vulnerability by asking for support has shown to increase overall acceptance by endearing the robot to the user. Indicating that these traits, which by our definition would be considered needy, are popular among elderly users.
Expression. Expressions, both visual and mental, are an important aspect of viewing SARs as a potential companion, rather than a simple tool, and are therefore important for creating a robot that users want to care for. Showing basic human-like characteristics can help with allowing users to feel that they can understand and relate to a SAR. However, these expressions do not always need to be a direct mirror of human expression. Subtle cues that portray emotions can allow users to empathise, even in instances where the SAR is not a humanoid. In order to be perceived as natural these expressions need to be manifested immediately during communication.
Personalisation. SARs should be tailored to suit the preferences of individuals, ensuring that their requirements are kept at the forefront of design to encourage prolonged adoption. Robust personalisation also allows SARS to be adapted to support users with disabilities, increasing overall usability by making sure that vulnerable users are not excluded. Personalizing less functionally significant aspects, such as the voice, can help to appeal to different demographics and increase overall enjoyment.
Understanding. Voice control integration has proven to reduce the learning curve for elderly users , however, barriers to using this method of interaction still need to be considered. While speech recognition technology has seen significant advances, concerns remain surrounding how well they can understand user requests. Elderly people in particular are hindered by this as speech may be slurred or otherwise difficult for the device to understand. This is compounded by a lack of support for multiple languages, as individuals would rather communicate through their first language, which may not be possible. Users being unable to understand existing SARs is a common source of frustration. Enunciation from speech-based robots needs to be clear, as the elderly are more likely to have hearing disabilities. Additional methods of communication, such as through a display, can lessen the severity of these issues.
READ more on this here: https://link.springer.com/chapter/10.1007/978-3-031-42293-5_42
