AI for persons with Disabilities

Introduction

Artificial Intelligence is the ability of a machine to simulate the human mind. In these times, the current generation is actively trying to ease their burden and shift the workload to the machines. Its relevance applies to the whole kit and caboodle. From industrial to entertainment, all sectors are crucially dependent on AI. With the ever-growing need and dependency on AI, the roots of the same can be found deep in the healthcare sector too. Current strides in the sector involve inculcation of AI powered tools to perform precision surgeries, monitoring the health status of the patients and drug discovery; to name a few.

In this article, the focus will be on the role that AI plays in the advancement of life of disabled people. The How’s, Whys, and the What’s. This article also aims to explore a few case studies to judge the current subservience of the technology.

What is a ‘Disability’? How has AI helped so far?

According to UPIAS (Union of Physically Impaired Against Segregation), disability has been defined not as an impairment of the body or brain, but as a “relationship between people with impairment and a discriminatory society.” Abiding by the definition, we can say that the main task for AI in this domain is to bridge the lacuna between society and the people with impairment. The goal is to create a society which believes in equity while transcending the barriers of bias.

According to the statistics, more than a billion people in the world are disabled. So far, there have been a plethora of developments to assist the disabled. We can classify the areas where the disabled face issues into 4 main sub domains:

• Communication with others and being connected: Communication is a key skill for all humans to sustain themselves. Since humans are so inter-dependent, they require basic skill to communicate their various needs to function in our day to day lives. Unfortunately, people with different kinds of impairments struggle with the most basic of tasks. This is where AI has stepped in to make the lives easier for this segment of the populace among others. There are voice assistants like Siri, Alexa, Cortana which make it convenient for blind or visually impaired people to communicate their requirements. Other technologies include Readjustments of graphic elements such as fonts, colours and spacing for people with a visual impairment. Audio descriptions content for people with a visual impairment etc.
• To get around: Mobility is another important aspect of our daily lives. Disabled persons must prepare extensively to even travel the smallest of distance. Things like wheelchair ramps and places with easy parking spaces, etc. are to be considered. For this purpose, applications like Moovit, Google Maps etc. have been developed with special features which highlight places and areas that have special provisions to include people with various disabilities.
• Independent Living: Integration of home objects with the Voice assistants was one of the many commendable steps which the various tech giants focused on introducing as a part of our day to day lives. Having a well-connected smart home can sometimes prove to be lifesaving. If a person falls, they can call the emergency services via the voice assistant tools they set up.
• Accessibility to same services as everyone: Inclusivity means that all the people have right to access all and any services, irrespective of their disability or profile. The shortage of Braille tutors was substituted by AI, and the gamification of the learning process enabled the children with impairment learn efficiently and have a better understanding. It was a laudable attempt to make the experience of education for these children as close to real and normal as possible.

Case Study

So far, we surmise that the AI has been utilitarian. The prevailing use cases depict that the technology is reliable. However, there are a plethora of shortcomings when it comes to making it autonomous. However, the next step is to aim to remove the human intervention entirely. With the help of this use case, a better judgement of the current scenario will be presented. The accounts have been summarized, and in the form of a diary entry.

Case of Peter

Peter, a man living in UK, fell down the staircase while returning to his bedroom in the dark, early in the morning at 5. He broke his neck in the incident and it was diagnosed that he had injured his spine. Because of the injury, he lost his ability to walk, speak and eat entirely. After 6 months of Physiotherapy, he recovered, but now is handicapped.

The aftermath of the incident affected Peter greatly. His day-to-day accounts are as follows:

7 AM. My carer wakes me up. He moves me and watches over me in the bed during the night. The caretaker serves me a light breakfast, my morning medication. A second caretaker then arrives to prepare me for the day and wash me up.

9 AM. Then, my carer gives me a wash and settles me in. I check my medication with my caretaker, and then order some pain meds, so I use speech activated software to contact my local drug store. The process is a bit cumbersome. After a few tries I can call the place and order the medication. The call is answered by no one, unfortunately. I can even hang up the call. Then I am expected to leave a message, which I do not really want to do, but I cannot stop the call either. My carer arrives from the kitchen after prepping the meal and switches the phone off for me. After another try, I am successfully able to place a call.

10 AM. I start my work and decide to answer all my pending emails. For this, I use speech recognition and the assistance of a carer. The carer helps me with the physical tasks involving the computer setup. He switches on the devices, logs in, and enters my e-mail information. When I was in hospital, I was trained to use a device which utilized a camera. It recognized a silver dot placed on my forehead or on my pair of glasses. Unfortunately, the software, although ambitious, was extremely annoying and difficult to use. I then started asking my carer to operate the mouse for me. I use speech technology to help me answer my emails. The technology is trained to recognise my voice and most of the time it does okay; however, on several occasions, it spells words incorrectly, and sometimes fails altogether. In the time it has taken me to type and correct this paragraph (including asking my carer to retype several of the words), I could previously have typed two pages. Simple commands which involve words such as “Word” are likely to open the word processing software “Word”. As another example, asking the speech software to type the word “dot” resulted in a “full stop” being typed. In many ways the speech software is a tremendous help; in others it is very frustrating.

2 PM. I have a Zoom meeting with two former colleagues. My caretakers manage to launch the meeting for me successfully. However, after about 20 min, I must have said a phrase which resembled the word “bring up”. This executed the dialogue box command and it stayed on the screen. Although my colleagues cannot see it, I can no longer see them. I need to shout the command “go to sleep” at my screen. Sometimes this works, sometimes it does not. More than often, I must depend on my caretaker who closes the dictation box and switches off my microphone which operates the speech software. Another minor irritation over.

9 PM. I have a snack, the caretaker gives me my evening medication and I tell my home hub to “go to sleep” which takes several attempts as it often ignores me, and we settle down to watch some television. Another day over. A similar routine begins the next morning.

Analysis of the Case

The case study shows us that although Artificial Intelligence is conducive, but we still have a long way to go. For Peter, the User Interface was designed to be as convenient as possible. But unfortunately, even after all the efforts, the User Experience was not pleasant. To aid the disabled even more efficiently, there are still a lot of areas of improvement. For instance, the speech recognition software. It was not as robust and not as accommodating as Peter wanted it to be. Ultimately, he had no choice but to take aid of his caretaker. The voice commands were faulty and were not entirely accurate.

All these points give us a new conclusion: AI is not at its peak, as of now.

Guidelines for an Ideal AI System

Accenture, the tech giant, laid out 4 basic principles in their paper ‘AI for Disability Inclusion: Enabling Change with Advanced Technology.’ It interprets the foundational rules, any organisation should follow to make AI ideal for the disabled users as well.

• Responsible: Any organisation should adapt and scale AI responsibly and should ensure complete transparency, compliance, and accountability. Is the technology affecting any group negatively? How is the AI affecting the lifecycles of the people involved?
• Accessibility: The organisation should prioritize accessibility to the clients including the feature and functionality tools as well. The company should ensure to embed all privacy and accessibility policies.
• Inclusive: It should be kept in mind that the organisation should act in fairness with AI. Using inclusive design approaches that also incorporate the disables is of utmost importance. The AI should not be biased.
• Secure: The AI should never put the privacy at risk. The individuals need not ask for help just because of their disability. It should be noted that the disabled will have the same expectations from the technology, like any other individual.

Conclusion

There is a lot of activity in the field of AI and Technology. It is commendable the way it is advancing day by day. Sometimes AI lives up to its expectations. Sometimes it doesn’t. The overall result of the efforts remains positive. The trends and the feedback inform us of the areas where the “advanced” technology fails. Some of the menial human tasks are still difficult to accomplish for the disabled, even after the aid of modern paraphernalia.

There is still a lot of room for improvement. Soon, the AI is expected to be the perfect companion not just to the disabled, but to the entire mankind.

References

• Artificial intelligence and disability: too much promise, yet too little substance? (2020). Springer.com. https://link.springer.com/article/10.1007/s43681-020-00004-5