- February 7, 2024
- Artificial Intelligence (AI)
How would you feel about a robot or AI software caring for you in your old age; helping you eat, move, clean yourself, or simply keep you entertained? This is starting to become a reality. Many countries like Japan are seeing a concerning shortage of human caregivers for elderly and disabled people; 63% of care homes reported a shortage of staff in 2022. Analysts have predicted that by 2040 in Japan, 690,000 additional carers will be required to meet demand. As a result, many varieties of AI and robot caregivers are in development to help tackle this; ranging from physical care assistants to emotional support bots.
However, amid this innovation lies a delicate balance. While these technologies hold promise, studies have shown that some existing robots have inadvertently added to the workload of human caregivers. Moreover, as robots and AI assume caregiving roles, profound questions arise concerning safety and ethics. It’s crucial to scrutinize the impact of these advancements on the well-being of both caregivers and care recipients.
Physical Caregivers
There are many examples of physical robotic caregivers in fiction. Think Baymax, the personal healthcare companion from ‘Big Hero 6’, or Markus from ‘Detroit: Become Human’. These are extremely advanced versions, but that’s not to say they’re impossible. Similar ideas are in development, although developing highly sophisticated robot caregivers with capabilities akin to humans is a complex task. It could take many years, if not decades to achieve.
‘The Hug’
Most current robots are semi-autonomous at best, meaning they need someone present to control or supervise, move, and maintain them. For example, Fuji Corporation recently invented a mobility support robot named ‘The Hug’ which can assist caregivers in moving people. The idea was that fewer caregivers would be required to move patients, also reducing the physical strain on them.
However, while this robot is reportedly easy to use, it doesn’t incorporate AI and relies on people to control it. Additionally, research carried out by Dr. James Wright uncovered a strikingly negative response to the Hug from staff in care homes. He reported that while the physical strain on the staff was reduced, the time taken to move patients significantly increased and many patients found the Hug uncomfortable to use. Consequently, they decided to stop using it.
Assistive Feeding
The feedback regarding The Hug indicates that for robots to help and complement the work of human caregivers, at least some AI needs to be incorporated into them. An example of an AI-based robot that’s currently in development is an assistive-feeding robot consisting of two robotic arms. Robots for this purpose already exist, but they typically make pre-programmed movements. They also must be specifically set up for each person and meal, they generally only lift the food to the mouth rather than into it, and they usually require someone to control them.
However, this new assistive-feeding robot uses several algorithms to autonomously carry out the process for a wide variety of foods. Most of these algorithms use computer vision. They visually determine where the food is, where the person’s mouth is, and when they have taken a bite. The goal is for the movements to feel as human and natural as possible. One of the main challenges is ensuring that the device can handle all food types, especially soups and fragile items. Nevertheless, the progress that the team has made looks promising for the future.
Other Examples of Physical Caregiving Robots
Other successful and well-received physical caregiver robots include those that help people bathe, as well as health monitors that learn the individual’s “normal” levels and can send alerts when unusual patterns or movements occur. While these devices don’t remove the need for human caregivers, they do reduce the attention each person needs from them. This allows caregivers to spread their services between more people.
Safety Concerns for Physical Caregiving Robots
For the future of physical caregiving robots, arguably one of the most significant challenges is ensuring their safety. As fragile lives are in their hands, a simple malfunction could be critical. Some safety considerations are:
- Accidental harm: Elderly people are particularly at risk of injury. Even human caregivers can find it challenging to deliver the appropriate degree of tenderness. Therefore, significant work needs to go into ensuring that robots can handle people as gently as possible, without bumping or pinching them, or colliding with objects. Sensors need to be present that can detect obstacles. Additionally, the materials that the robot is made from must be carefully selected.
- Emergency situations: Robots should be capable of handling emergencies if they occur, such as falls or sudden deterioration of the individual’s health. This may involve alerting caregivers or emergency services, or potentially even performing first aid in particularly advanced future robots. However, learning to trust a robot in an emergency could be difficult for some people.
- Monitoring and compliance with regulations: Continuous monitoring and upkeep of the robots are essential to avoid breakages and malfunctions as much as possible. Additionally, adherence to safety standards and protocols established for these types of devices is important.
- Adaptability: Meeting the diverse needs of disabled and elderly individuals is a complex task, given the vast spectrum of health conditions and requirements they may have. Ensuring that a robot caregiver can effectively cater to this array of needs poses a significant challenge, one with seemingly boundless possibilities. However, machine learning algorithms offer a promising solution by enabling robots to learn and adapt to each person’s unique behaviours and preferences over time. Coupled with computer vision and camera technology, these robots can make real-time adjustments to their caregiving approach, ensuring effective support in any situation.
Social, Mental and Emotional Caregivers
Physical conditions and lack of mobility aren’t the only factors making life difficult for older people. A report from the National Academies of Sciences, Engineering and Medicine revealed that nearly a quarter of adults aged 65 and older in the US are considered to be socially isolated. Social isolation and loneliness can impact their health in multiple ways:
- Can cause depression, anxiety and suicide.
- Associated with approximately a 50% increased risk of dementia.
- Also associated with a 29% increased risk of heart disease and a 32% increased risk of stroke.
For these people, the direction AI-powered devices are taking can literally save their lives and could also prevent the need for them to move into a care home if they are physically able to take care of themselves. Fortunately, engineers are rapidly advancing robot or AI caregivers in this area, outpacing developments in physical care. These robotic solutions can be relatively straightforward, or they can leverage existing technologies such as mobile phones and computers. Even a basic AI-based mobile app can yield a remarkable improvement in an individual’s mental well-being.
In recent years, there have been significant advancements in Natural Language Processing and Generation. Chatbots like ChatGPT understand even the most complex text and give intelligent, human-like responses. Additionally, computer vision models to identify human emotions from faces and to identify objects in an environment are becoming increasingly common, both regularly achieving accurate results. The progression in these areas combined unleashes a great deal of potential for the functionality of social robots, which can allow them to start relevant conversations based on what’s happening around them.
Kismet
Dr. Cynthia Breazeal designed and developed Kismet as one of the first social robots for her doctoral studies at MIT in the 1990s. Its primary function is relatively straightforward: recognising and simulating human emotions, facilitating face-to-face interactions. While it cannot comprehend and generate human speech, Kismet represented a significant advancement in its era, laying the groundwork for the social robots we see today.
Jibo
A more recent example of Dr. Cynthia Breazeal’s work was Jibo, a type of personal assistant robot marketed towards families. It has facial recognition capabilities, so it can recognise different family members and it can understand and respond to human speech. It also has various apps called ‘skills’ that allow it to entertain, tell stories, dance, play games and take photos. Acting as a companion to socially isolated people, it could make all the difference to their mental health, allowing them to engage in conversations whenever they like, and keep them entertained.
Aibo
Pets can have a significant positive influence on their owners’ mental health. However, not everyone can own a pet, and in the case of elderly or disabled people, they may be unable to provide the required levels of care for the animal. This is where the idea of robotic animals came from, with findings suggesting that they may be promising in relieving depression and improving quality of life, despite their lack of vocal communication.
‘Aibo’ is a robotic, AI-powered puppy developed by Sony. It’s intended to be a lovable companion that develops its own unique personality over time. Like a real dog, over time it will get to know the name that you have given it, learn tricks, play, and get to know members of the household. Additionally, Aibo can return to the charging station by itself when low in battery, perfect for people with disabilities.
Concerns for Social Caregiving Robots
Like the physical caregiving robots, there are also some concerns regarding social caregiving robots. These may include:
- Unhealthy attachment: The human-like companionship and support provided by the robot could result in unidirectional emotional bonds, potential for deception, and subconscious influence of the robot on the person. It could cause these people to become overly reliant on their devices, choosing them over real human connections. People with cognitive impairment, like dementia, may be particularly at risk of this.
- Emergency situations: Emergencies regarding mental health can be even more complicated and difficult to handle than physical emergencies. AI and robots must be capable of handling a wide range of sensitive issues. This is an ongoing issue with NLP and NLG models – there are many records of people reaching out to AI for help regarding very serious and sensitive topics and receiving inappropriate responses.
- Privacy concerns: As large quantities of image and text-based data is processed and stored, it must be appropriately handled and encrypted to protect the privacy of the user.
Conclusion
The growing field of AI and robot caregiving holds significant promise in addressing the growing demands for elderly and disabled care. However, as highlighted, realising this potential requires careful consideration of the complex challenges and ethical implications inherent in deploying such technologies. While physical caregiving robots offer tangible benefits in alleviating the strain on human caregivers, ensuring their safety remains paramount.
Likewise, social caregiving robots present opportunities to combat social isolation but must navigate their own concerns regarding the fragility of mental health and privacy. By addressing these challenges proactively, we can harness the potential of AI and robotics to enhance the quality of life for individuals in need of care.
Discover what we can offer you at Vsio Applied Analytics by viewing our Services, Case Studies, or Contact Us today!
