Healthcare Tech

How Artificial Intelligence is Remarkably Improving Assistive Technologies 


How Artificial Intelligence is Remarkably Improving Assistive Technologies 

As humans, we have some basic needs. We perform activities to meet these basic demands of food, water, and shelter. Everyday activities such as eating, dressing, etc., are an essential part of living, but even these simple activities may sometimes be challenging to perform for the elderly and disabled individuals. This is when it becomes difficult for them to carry out these basic tasks without a human touch of assistance.  


Many advanced assistance devices, such as assistive robots, have been developed to help individuals with disabilities perform daily tasks independently. These tasks can, however, be difficult if they involve both coarse reaching movements and fine-grained manipulation. For example, switching between food items while eating is made easy, by granting full control over the food. 


It isn’t fair for an old person or an individual with disabilities to be deprived of a real dining experience. If the caregiver is unavailable, the individual would not be able to feed themselves in the former’s absence. The resultant feeling of shame that would envelop the person would probably make them ask for smaller amounts of food to shorten the eating time or even reject meals altogether. Assistive technology, sadly, is not always readily available today. Instead of assistive robots, a treatment plan is issued for those in need, wherein a caregiver will feed the client. While some do not have an apparent problem with this method of feeding themselves, others believe that assistive feeding takes away their right to feed independently. There are two reasons for not making this technology available: knowledge of various equipment and funding access. 

AI, assistive technologies, old person, long-term care, care, nursing, meal

Researchers have also come up with various assistive feeding devices for the physically handicapped using feedback control mechanisms. Some assistive feeding technologies include manual aids, robotic technologies, and electrical advancements. Manual devices enable the user to self-feed with the help of a non-electrical aid. These include adapted cutlery, grasping aids, height-adjustable tables, spring/hydraulic feeders, and manual mobile arm supports. In contrast, electrical and robotic aid includes battery-operated cutlery, power-assisted feeding, and computer-programmed robotic feeders and arms. Power-assisted devices are primarily operated by switches or a joystick to control their components. These devices come in handy for users with uncontrolled movements, severe weakness, or rapid loss of strength in their upper limbs.  


Over 1.5 million Americans rely on wheelchair mounted robotic arms to perform routine activities. These arms are difficult to control and take a longer time to respond. Stanford researchers have come up with a unique and faster algorithm to control assistive robotic arms. Experiments have proven that this controller allows individuals to cut, shovel, pick, scoop, and sprinkle food items. Hong Jun Jeon, Dylan P. Losey, and Dorsa Sadigh developed a controller to blend two artificial intelligence algorithms. The first allows two-dimensional control of the joystick on the robotic arm. It uses contextual clues to assess information, for example, whether a user is trying to reach a doorknob or a drinking cup. Then, once the robot arm is approaching its target, the second algorithm comes in to allow for more accurate gestures. This control is shared between the human and the robot hence enabling communication between the two. 


Among the challenges associated with eating is the need to cook and move food, a capability that assistive robotics has sought to make feasible. Researchers have developed robotic algorithms that can individually manipulate and distribute various food items to people with disabilities. However, the design of a fully autonomous method to manage a task as complex as feeding is really challenging. It is rightly said, “one size doesn’t fit all!” 


Shared autonomy can make it easier to eat by anticipating individual intentions and then increasing the robot’s input. But current practice focuses on goals such as helping the human move the robot between some food options. The robot may use demonstrations to keep track of the map between the person’s low-dimensional inputs and the high-dimensional behavior of the robot. This approach is feasible when moving along a line of preferences, but in the case of noisy human inputs, the robot will tend to move away from the goal. 


Technology may allow someone to live with more freedom and liberty if used properly. It can also encourage family members and friends to become more involved in the care of a loved one. The use of assistive technologies for elderly or disabled individuals might be seen as a step towards total independence from caregivers, but this is a wrong approach. You can look forward to artificial intelligence as a helping hand. Assistive feeding technologies theoretically minimize the care package’s expense, provide the individual with independence, and improve physical and mental well-being.  

AI, assistive technologies, old person, long-term care, care, nursing, handicapped, robot, selfcare

Let us have a look at some of the major advantages of assistive technology: 

  • Assistive feeding devices increase self-care 
  • Boosts self-esteem with a sense of independence 
  • Increased safety during meal-time 
  • Freedom for caregivers to carry out other chores 
  • Meals made better and satisfactory for individuals 

The usage and development of emerging technologies for the disabled and elderly is vital, as it is not simply a matter for these individuals to do the routine tasks in an easier way with the help of a tool. But it is a matter for them to carry out certain tasks individually and, perhaps, to learn how to carry out new tasks to improve their own autonomy. A self-feeding robot may be helpful in the individual’s view if the user sits alone at home. The robot can be seen in two situations: one is self-feeding, and the other is eating together in a group. The most critical task is to encourage self-feeding without a caregiver as people with disabilities remain alone in their homes, and caregivers cook food in advance. The person should be able to adjust the speed of the arm.  


Robotic arms with assistive feeding technology are a result of several studies. Some challenges come forth while implementing the latest assistive technologies for daily tasks. The first challenge is to teach the robot how to poke food. It may sound easy, but imagine the difference between how it feels to pierce the surface of a crunchy carrot, a squishy grape, or a soft banana. According to Gilwoo Lee, “People change the way they skewer food based on food shape, hardness, ease of feeding, etc.” If you pierce a banana from the tip, you’re going to get through the skin easily. But as you try to pick it off the plate, the slippery banana will roll straight down the fork. This is a motivation to develop an angled skewering motion, which significantly increases the success rate for eating soft items. The next challenging task is to bring the food to the right height and hold it onto the fork. These small but intriguing challenges will shape the future of AI and assistive feeding in the coming times. 


Artificial intelligence can be a game-changer for people with disabilities. The world population is over 7 billion, and over 15 percent of the population is physically impaired. According to surveys, just one in ten people with a form of disability can access some form of assistive technology or goods. Future work may be the combination or convergence of various forms of bio-signals to provide additional contact and control possibilities. The occurrence of the key signal being influenced by the other signals should be considered. Besides, studies are being conducted to modify and update the assistive feeding devices to offer more comfortable and quicker eating. 


Previous Article

Dermatology has a race problem — AI might make it worse

Next Article

How AI is Used to Detect Deep-Fake Videos

Related Posts