2012 was a big year in Great Britain. The country was awash with national pride. Union Jacks proudly flying at full mast across most British streets and a feverish party atmosphere, as we hosted arguably the best ever Olympic Games. During the games, we witnessed the prominent rise and rightful public recognition of the athleticism of disabled athletes, competing in the Paralympics and Special Olympics. Tanni Grey-Thompson, Ellie Simmonds and Lee Pearson are now just as likely to be debated as Britain’s best Olympians from those games, alongside Mo Farah and Bradley Wiggins, amongst friends down the pub. Whatever stigma previously existed was smashed to bits, as our heroic athletes brought home gold medal after gold medal.
As a result of the games, we are now seeing society becoming far more
understanding and inclusive of the disabled community, but with the United
Nations reporting that around 15% of the world’s population suffer with some
type of disability, there is a growing feeling that we need to do more to help
make the lives of people with disabilities easier.
Data Science and Artificial Intelligence has come to the forefront of
technology in the last few years, and several practitioners are taking a more
philanthropic outlook on life, supporting people suffering with both physical
and mental disabilities.
One of the areas where Machine Learning is playing a prominent role is the
support of people suffering with Autism Spectrum Disorder (ASD), which is a
condition suffered by approximately 1 in every 100 people, with men more likely
to be diagnosed with the condition than women. It affects children at around
age 3, and results in the child having difficulty processing or engaging in
human interaction or emotion, which can make integrating into groups of other
children very difficult. However, the London Knowledge Lab was incredibly
successful in introducing a group of ASD children to a virtual autonomous robot
called Andy. The experiment found that the children interacted readily with
Andy, listening intently, asking and answering questions far more freely than
if it had been a human adult. Similarly, an artificial intelligence robot
called Milo has been rolled out to over 50 schools in the United States to
encourage children with ASD to interact more readily face to face, and in some
cases, even permit physical contact from the robot, unthinkable for the
majority of ASD sufferers. While there is no recognized cure for ASD at present,
the experiment offers hope to the families of children and adults suffering
with ASD, that they might one day be so comfortable conversing with artificial
intelligence software that they grow in confidence in their interactions
with other people.
Intelligent prosthetics is another area in which Data Science has been
implemented in order to make the lives of disabled people easier. For many
amputees or people living without the use of one or more limbs, the options are
very limited, and often very rudimentary, focussing on bare essentials rather
than everyday usability. However, Deka Research, headed up by Segway inventor
Dean Kamen, have pioneered a prosthesis they call the ‘Luke’ arm (named in
tribute after Luke Skywalker). The unique aspect of this piece of technology is
that its modeled more closely on the internal workings of a human arm, with
the mechanical equivalent of tendons, muscles and bones allowing the user a
much more natural range of motion, closer to a human arm than to the hooks and
claws that have been more prevalent in the last 20 years. The arm itself can be
controlled in several ways: with microscopic nerve endings attached to the base
of the arm, or controllers in the wearer’s shoes. The project was granted a
significant financial boost by the US Army Research Office, but the prostheses
are likely to still be incredibly expensive to produce on a mass scale, with
some estimates at around the $50,000 mark. While this may not be the answer for
everyone who needs a prosthetic, the range of movement and sophistication of
technology on show here is a brilliant sign for the future, when similar
technologies will be available for less cost, and potentially with even more
uses.
A piece of technology that was brought to prominence by one of the world’s most
famous scientists, Professor Stephen Hawking, the electronic augmentative and
alternative communication system (AAC) has become integral to the lives of
those people who are unable to speak due to conditions such as motor neurone
disease or cerebral palsy. One of the latest innovations in this market, and
still closely related to the ‘Equaliser’ device used by Hawking from 1986
onwards, is the DynaVox EyeMax. This device uses Computer Vision techniques,
via a front-facing camera to track the movements of the user’s eyes across the
screen of commands, and can even be programmed to use certain intonation in
order for the speech to not be misconstrued. It has the potential to use
Natural Language Processing to give speech to a large number of people who may
have lost the ability, but there are some limitations to the technology; each
device must be individually programmed for the respective user, to include
places of interest, names of friends and family and other unique information,
and the cost of the devices is not often covered by health insurance. It is, however, expected that the inclusion of deep learning technology into the
devices in future should reduce the time taken to program them, which
could bring the cost of the devices down to a more affordable level.
It is not only specialist medical firms who have seen the value, both financially and philanthropically, of using data science technologies to improve the lives of disabled people. Google recently announced that they were looking to boost the accessibility of their smartphones and other devices for users who might have limited or no use of their sight, hearing or dexterity. One of the most exciting and unique developments is a Braille system for use with Google smartphones, which allows users to connect a Braille device to their phone via Bluetooth. In addition, Android systems can now be controlled by ‘switches’, not dissimilar to those used by Stephen Hawking in his wheelchair. This, according to Google, opens them up to a new group of users who previously would not have been able to access this technology, and allows users to use the same hardware as their friends rather than having something custom-made and potentially less appealing. There is a huge financial incentive to these developments, as both Switch Access and BrailleBack are free to download for Android, meaning that unlike many technologies designed for the disabled community, price does not have to be a barrier. Overall, Google’s contribution has laid down a benchmark for other technology companies to follow their lead in producing affordable and easy-to-use platforms for those with disabilities, and we look forward to see who steps up next.
Finally, there is something that is definitely more at the ‘proof of concept’ stage right now, but shows enormous potential for the future. In 2011 Dr Dennis Hong, from Virginia Tech University’s RoMeLa Robotics and Mechanisms Laboratory, pioneered a car that was designed for blind or partially-sighted people to be able to drive independently. Using a system that combined machine learning software that was able to learn a pre-determined route and sense obstacles and pedestrians, and a series of sensors packed into pressure-pad gloves, that would tell the driver where they should be going and if there was an impeding obstacle that needed to be avoided. During his TED talk in the same year, Dr Hong demonstrated the car with a blind person driving, doing a whole lap of a pre-determined course safely at the Daytona Speedway. It must be noted that there are a few reasons why this system could not be implemented immediately, least of which is that the route had to be meticulously entered into the car’s computer in order for the sensors to direct the driver, and this would not be feasible for every road. However, it shows enormous progress in the field of Data Science and Machine Learning, and a huge leap forward for the blind and partially sighted community as a whole.
There are several questions that must be addressed before the relationship between Data Science and Disability can be labelled a success. Firstly, are the technological innovations that are beginning to emerge at an advanced enough stage to benefit those who are suffering from a disability now, or is the real breakthrough likely to come in 5-10 years? Similarly, are these options all financially viable, not for the producer, but for the consumer, who might only have limited income as a result of their income? Furthermore, are innovations offered by Data Science limited to making the lives of disabled people easier, or is there scope, for example, for deep learning to be used to identify the genes that cause certain disabilities, giving doctors the chance to study and limit the genes in the future?
The future for Data Science looks incredibly bright, as several innovations are already making marked differences to the lives of disabled people around the world. The next step, whether that be into genetics research, deep learning integration or another field, is dependent on whether many of the major companies will look past the finance sheets and instead look to improve the lives of the many millions of disabled people all over the world. Let’s hope it doesn’t take till the next Olympic Games until they do.