Monday, 20 February 2017

Tech for helping Chronic Diseases

This week I'd like to focus on an amazing innovation that helps people with chronic diseases like Parkinson's, Huntington's, cerebral palsy and mobility issues like tremors or limited mobility. It is a special spoon produced by the company liftware that is aimed to increase the quality life of those suffering from postural tremors in their upper limbs.
Without further ado lets delve into its design and specifics.

IxD 


The interactive design for this device is really simple and effective. It is basically a spoon (or a fork) that allows you to eat without a lot of help. People with hand/arm tremors find basic day to day activities quite hard and it usually affects not just their quality of life but also their confidence. Think about going to a restaurant, eating soups and things like ramen that need precise movements even by those with practised hands. However for people with severe conditions of tremor this means extremely slow eating, eating with assistance and sometimes preferring not to eat certain kinds of food! Doing something as simple as eating without assistance is of incredible importance in such situations.
The Liftware website is also very specific in its instructions about who this device can help, what kind of tremors it can help with and to what degree. It provides a simple test to those who would like to know if they're suitable users for it and how much they recommend it usage. There also several videos that show how the spoon can be used and allows the potential buyers to view it being used by others in a similar situation as them.


UI


Liftware spoon is quite affordable for what it offers. According to its amazon price, a starter kit of Liftware is ~200 CAD which for the services it provides should be worth it. It is also aesthetically pleasing in a simple and clean design in white color. It also comes with its charger in a small box that can be conveniently carried around.
The liftware spoon works by counteracting the tremors through two motors that work opposite to the tremors and stabilize the hand by using motion sensors in it. In case of liftware spoon for people with arm mobility issues, the sensor detects and works against unintentional tilting or tipping. A small onboard computer distinguishes between the unintentional movements and intentional hand movement.
Since the device uses magnets to attach the utensil attachments, they recommend consulting with healthcare providers if the user has a cardiac pacemaker or other electrical implants devices in their
bodies, but other than that it is safe to use like a regular utensil.


UX


To provide a better experience, the Liftware spoons come in two different types that are directed for both people with tremors and those with hand and arm mobility limitations separately. Both are directed for their specific uses and work with sensor technology. It comes with 3 different utensil attachments of spoon, soup spoon and a fork.
It has a decent battery life of upto 1 hour of continuous use that could last for ~3 meals on a single charge.
It is even simple enough to clean by detaching the utensil and cleaning like regular ones, and the stabilizing handle can be cleaned with a sponge or disinfectant.


Other details


What makes this technology great that its made for some very specific uses that often taken for granted. They also have the research done with it to show an effectiveness of 70% less tremors that is available for users to look into and make intelligent choices.

Bibliography:

https://www.liftware.com/

http://onlinelibrary.wiley.com/doi/10.1002/mds.25796/abstract

https://www.amazon.com/Liftware-Steady-Starter-Kit/dp/B00JDSIOJE

https://www.youtube.com/watch?v=fS01kn6YJ94

Monday, 6 February 2017

Tech for the Visually Impaired: Keeping with the Theme

Trying to find a blog topic last time opened a whole world of options to me. However there were some innovations that just seemed too (for-lack-of-a-better-word) cool. So this week I'd like to build on tech advances made for the disabled by exploring the world of those visually impaired. 

Airpoly Vision is a free application made by Airpoly designed to help people with visual impairments to identify objects and colors in the environment. It is available for download on the Apple App Store.  Here is a link to their official video: https://www.youtube.com/watch?v=XMdct-5bERQ 
and supporting video : https://www.youtube.com/watch?v=h5g82YNmwmU


IxD 


It has a really simple, easy to use and convenient design. Consider times when we have easily decided what to wear to colour coordinate with the perfect shoes, or walked into an office knowing we're walking into the right room or something as simple as knowing what's on your plate. While all these tasks seem ridiculously simple to those with proper vision, you can see the hardships involved in impaired vision. This app allows for users to overcome some very basic hardships. Asking for directions, reading road signs and even shopping alone is made easier. You only need to point the screen in the direction of/ at whatever needs to be labelled and take a picture for it to provide a speech or text of it. This allows people with visibility problems to visualize the space and things around them.


UI


The app has a simple and aesthetically pleasing look. How it works is, once the picture has been captured it gets uploaded to a server that uses a system of convolutional neural network to look through the pictures. It basically breaks it down to different points of interest that can be matched to a particular object. It not only provides an accurate name for the objects its given but also the description in terms of its color and its verb forms that show it being used example, "Man riding bicycle." 


UX


There are other apps in the market that target this particular group, that rely mostly on crowdsourcing which works for the most part with extremely good accuracy, but has the downside of slow timing and not to mention the need for an internet connection. Airpoly on the other hand, works without an internet connection making it accessible at all times and even handy for privacy. 

The user experience in general seems to be great. The app uses multiple language like English, Spanish, Portuguese etc to make it largely convenient and also allows other users to build on its knowledge of objects. 
Another exciting feature of this app is the involvement of users. Its not the same as feedback after since it is just going to update its versions with more options and accuracy but anyhow, you can change its labeling if you know its wrong or not completely accurate. This is of course both extremely useful and also a little ironic since the target audience is the visually impaired who would be the ones needing the assistance and not the ones providing clarifications. Nonetheless this particular function would be useful to improve its accuracy as it has more and more items in its database.

Airpoly vision quite accurately recognizes plants, animals, colour and even currency. While USD is the only currency in its database right now, it is working on adding more. It has some other convenient functions like detecting darkness, it automatically switches on the torch in your phone and once the light is back on again, switches it off. 

The app is also extremely sensitive with labeling a person as either man or woman and usually takes more time. It also has the ability to describe emotions at least the most visually visible ones like anger and happiness. However once again it takes considerable time before it calls someone angry. 

                                           

             

                                                            



Airpoly vision does not work in real time however, its best speeds are as less than 3-5 seconds with common day-to-day objects. According to their founders that is the one utility that they consider a true weakness. Its accuracies with objects seem to be increasing the more people use it. Another issue that is solved with more and more people using and correcting is misidentification. Another small issue is that you do need to direct it multiple times for smaller objects as it takes the background and surroundings into effect. So it might go "table..plate..food" to identify the food.  


Apart from technological features there's a few other things I'd like to mention. Any sort of artificial intelligence always raises ethical questions about who can modify and use it and how much. One has to wonder who controls for the changes made by users as it is plausible to think of malicious users trying to affect its accuracy for competitive reasons. What's stopping a user from mislabelling things and considering this particular app is aimed at the visually impaired, a wrongful labeling could be, in the worst case, a matter of life and death. A signal colour being misidentified as green instead of red or construction warning signs being misidentified can be extremely dangerous. 

I was unable to find it policies on user interference but I'd like to hope that like with any other AI technology, even Airpoly vision takes into account its ethical and moral dilemmas. 
As with any tech innovation designed mainly for the disabled, I'd like to hope that feedback and considerations will be taken from visually impaired people.

In conclusion, this app seems to work great and I look forward to using it on Android as soon as its released.


Bibliography


https://techcrunch.com/2015/08/17/aipoly-puts-machine-vision-in-the-hands-of-the-visually-impaired/


https://coolblindtech.com/aipoly-vision-artificial-intelligence-for-your-ioss-camera/


http://tech.aipoly.com/


https://itunes.apple.com/us/app/aipoly-vision-sight-for-blind/id1069166437?mt=8