Market Research

Forum Responses

As part of my market research I posted a rough proposal of my idea on public forums. The forums that I posted on were http://www.alldeaf.com and https://www.actiononhearingloss.org.uk.

Proposal:

Both Virtual and Augmented Reality are becoming more and more popular in today’s society.  But for what purpose? The expansion of these technologies is being idolised for all the wrong reasons; they are being incorporated into mobile applications and video games for people to enjoy and have fun with. I believe that we should use these technologies to our advantages and therefore I wish to create a pair of smart-glasses that will act as an assistive device for those with hearing difficulties.

The smart-glasses will incorporate Augmented Reality. The AR function enables the user to see a visual of a human being (man or woman) performing sign language superimposed in the real world. The microphones built into the frame of the glasses allow spoken dialogue to be recorded, the software inside the glasses will read the recorded audio and quickly translate it into British Sign Language. (To begin with, the only language available on this device will be BSL, however there will be room to develop the glasses so that alternative languages, such as ASL, can be added.)

Some may argue that one of the biggest difficulties of being deaf is that people of the hearing culture treat deaf people as if they are handicapped, people to be pitied or changed. Therefore, the design of the product will look like an ordinary pair of glasses. The main reason for this is to prevent discrimination from people of the hearing culture. However, this design will have further benefits, such as being comfortable and stylish.

It is apparent that a fairly large percentage of people who are deaf (especially those deaf from birth) would opt out of the hearing world altogether – no matter what technologies were offered. These smart-glasses are not a solution for curing deafness, they are a device which can help those with hearing difficulties to overcome issues that occur naturally as a consequent of hearing loss. The smart-glasses will hopefully be beneficial to those who cannot or struggle to lip read. The smart-glasses will also enable users to become more aware of surrounding conversations and may open the opportunity for further communication. The smart-glasses will also be suitable for children, particularly school children who may struggle to understand their teacher or peers.

Although there are many mobile applications already available that offer sign language translation assistance, there is no product as such currently on the market. Furthermore, this product has additional benefits because it is eyewear; prescription lenses can be made available for the glasses if desired.

At first I didn’t get any responses but after time a few people began to voice their opinions. One of the main concerns was that not many people know sign language and therefore a speech to voice option may be more beneficial.

“Text technology has freed up many with hearing loss, I just feel sign does the opposite and is restrictive. Certainly sign is no opposition in media at all, given a straight choice, few would choose sign language as a media format of access these days. Not the 9m HoH anyway… I don’t believe media shares the access demand view of BSL. ‘888’ has killed sign access as a necessary media access form. It isn’t a matter of choice/preference or anything else, just it is a superior access format for the majority. So I don’t really follow there is any demand for everyone to learn it or for applications to provide it, the market must be extremely small for it. The advantage of text access is it allows better intro to mainstream and its grammar, I never saw the point learning an alternative that would provide issues of access when I want to read something or get a job, with a loss situation that provides barriers in itself, who needs it ?

The glasses thing has been tried in cinemas etc not really effectively, again subtitling was superior there too. You can be sure once we get instant text access to speech via some really easy means, the sign fraternity will demand a signed version. But only 5% of deaf rely solely on BSL… I believe once a really effective and cheap way is found to access face to face text via unobtrusive translation without a middle person, it will be taken up by us all. At that point, it will be a real game changer for the BSL deaf. Their peers will have real choices…”

– MervynJames4

 

“There are more people who are hard of hearing or gone deaf.. and all in between .. who lost their hearing later in life and therfore do not sign. There are probably 6-9 million of us. Therfore BSL would be underused.

It would be much more inclusive and the uptake would be greater if speech to text were used, without the need for a third person.. although there are apps for this out there…they are still not accurate enough to say follow a film or the news in real time. Now that would be something if this could be achieved.”

– Tiptop

Another concern was the cost of the project:

“It sounds like a great idea but I have a few reservations. First of all it will cost a lot of money to develop this. There are already BSL translation applications on mobiles and there is a danger that other more portable apps will come along while you are developing this. So you would be dependent on earning enough money from the finished product and I have doubts about that.”

– Cornishhandy

It seemed that there were a lot of negative responses regarding the project. So I decided to take on board what had been said and change the concept of the product slightly. The changes that will be made are based upon the response that BSL may not be a common method of communication for all deaf or hard of hearing people. The option to to have speech-text translation will now be incorporated into the design. There is reason to think that product will benefit from this change as the target market is likely to expand, people who do not know sign language may potentially be open to purchasing the product, as well as people who have slight hearing loss.

Sources: https://www.actiononhearingloss.org.uk/community/forums/deafness-and-hearing-loss.aspx?g=posts&t=14389

http://www.alldeaf.com/threads/new-smart-glasses-proposal.129168/#post-2509332

Budgeting

There are many factors to consider when proposing a budget for this kind of project.

The graph below shows a rough estimate for the hourly rate for chartered engineers who belong to the Engineering Council based on years experience.

Screen Shot 2017-04-15 at 22.40.05

Another factor to consider is location. The graph below shows that the hourly rate varies upon the location of the engineer with the UK.

Screen Shot 2017-04-15 at 22.44.25

Also, the hourly rate will vary depending on whether the engineer is self-employed or working for another company.

Screen Shot 2017-04-15 at 22.43.36

And obviously, the hourly rate will vary depending on the type of engineer.

Screen Shot 2017-04-15 at 22.44.12

Ideally the costs of the project need to be kept to a minimum, however, it is likely that in order for the project to have the best overall outcome the following roles would need to be filled:

Project Team:

Project Manager

Industrial Designer

Electrical Engineer

Mechanical Engineer

Software Engineer

 

In terms of wages, most employees will be paid based on experience. This could be anything from £15 – £60 an hour.

The project itself is realistically going to take between 1-2 years to complete.

The average amount of hours that one person works in one day is 8 hours.

If each Engineer/Designer (excluding the project manager) were to be paid £30 per hour for an 8 hour day of work then the total costs for just one day of work would be £960 (£240 to each employee).

It is unlikely that these employees will be needed throughout the whole course of the project.

The minimum timescale for the project is 12 months so based upon 3 months of work from each employee, working 8 hours a day, 4 days a week at the rate of £30 per hour the costs to hire these employees for 1 year will be around £46,080.

The wages of the Project Manager still need to be added on to this.

Assuming that the project manager will be paid roughly the same but for a period of 12 months, the costs for the overall wages of the above employees will estimate around £92,160. 

Other factors to consider:

  • Costs of Materials
    • Material to make the frame of the glasses
    • The lenses
    • The Augmented Reality Lens
    • The Software (Android)
    • The Microphone(s)
    • The Micro-projectors
  • Sales and Marketing
    • Promotional costs
  • Packaging
    • Costs of production
    • Costs of material
    • Costs of employees

Sources: http://www.payscale.com/research/UK/Certification=Chartered_Engineer%3a_CEng_(Engineering_Council)/Hourly_Rate#by_State

I understood that this project was going to be cost worthy but I didn’t understand just how much. I decided to get in touch with some companies via email to see if they could give me any advice on how to lower costs or even a rough estimate on how much they think the project is likely to cost overall. I emailed well known companies as well as freelancers. Only one person responded to my email and here is what he said:

Screen Shot 2017-04-15 at 22.52.32

James’ response got me thinking. I was unaware of how much a project like this would actually cost. I realised then that the costs of production for my project were going to be extortionate!

But then I figured that the costs of production may actually be a lot less because the plan is to only incorporate the essentials,  the elements that are in most smart glasses are not needed, for the purpose of this project is to create a pair of smart glasses that display an Augmented version of a woman singing BSL, there is no need for extra components such as navigation, headphones, bluetooth connection, wifi connection, photography etc. Hopefully, this will result in the overall costs being reduced.

Having said this, I’ve realised that I have not considered the costs of hiring a Sign Language Interpreter, and because a highly qualified professional interpreter is required, the costs may be expensive.

 

Sign Language Interpreters

Figures from the National Career Service website estimate that the hourly rate for a freelance sign language interpreter vary from £25 – £30 an hour. Wages for interpreters who are employed by another company can vary depending on experience and location.

Screen Shot 2017-04-17 at 11.39.17

https://nationalcareersservice.direct.gov.uk/job-profiles/british-sign-language-interpreter

I would need to consider the fact that additional charges may occur for hiring a sign language interpreter to work on a project like this, as they will not be performing their ordinary job role.

 

Summary of Expenses

I decided to make a rough spreadsheet of the estimated overall costs. My projected costs are based upon project tasks divided into five subject categories, Project Design, Project Development, Project Delivery, Project Management and Other Costs.

I have based the labour costs for employees (Project Manager, Industrial Designer, Software Engineer, Hardware Engineer, Mechanical and Electrical Engineers) on a wage of £30 per hour per 8 hour day. This may be subject to change. Therefore I have allocated an additional £7,680 to cover 384 hours of work.

The costs for the Sign Language Interpreter is based up 96 hours of work at an hourly rate of £50.

I have allocated £60,000 for all materials (including hardware and software).

The total contingency allowance is £17,100.00 to cover any overtime in hours, extra material costs, additional travel expenses and any other costs not mentioned.

The total costs estimated for the whole project is £129,430.00 based on a 1 year timeframe.

Screen Shot 2017-04-17 at 13.44.06

Investment

I understand that in order for the product to be successful and for the idea to be put into practice, I would need to consider funding. I believe that crowd funding is one of the best options. I am aware that companies such as Kickstarter are designed specifically for projects like this.

https://www.kickstarter.com

I could also consider contacting organisations such as the British Deaf Association or Action on Hearing Loss to see if they would offer partnership, sponsorship or any form of funding. I could also contact companies such as Mind Rockets Inc. who have existing products that serve similar purposes.

https://bda.org.uk

https://www.actiononhearingloss.org.uk

http://mindrocketsinc.com/#home

The Product

Branding

I began to think about aspects of branding with regards to my product that I am proposing. The first step I made was choosing a product name, I came up with the name “SYGN” which I created through an combination of the words “eye” and “sign”. I thought that this name was most relevant to the product.

My second stage was to begin thinking about a logo and the style of font that I wanted to use. Here are some examples of typefaces that I particularly like and think will work well in terms of branding. All of these fonts are available online at http://www.dafont.com

Screen Shot 2017-03-28 at 21.44.37

I decided that I preferred the Sans Serif style font as opposed the the Serif. So that narrowed my choice down to the top three fonts. I decided to chose the font Rodondo because this typeface was different to the others and I liked the curvaceous edges.

Logo Ideas 

Screen Shot 2017-03-29 at 16.30.24

For the logo I decided that I wanted to use an ear icon but I also thought about incorporating an eye as the product is eyewear. The image above is one of the final designs that I created on Illustrator.

Colour Schemes

Screen Shot 2017-03-29 at 17.16.28

In terms of colour schemes, I thought about using grey because it’s a neutral colour, beige (pink/brown) because it is often associated with skin tones and blue because I wanted a colour that would balance out the pink tones. I experimented with light and dark shades to see which ones I preferred. In the end I chose the darkest shade.

Screen Shot 2017-03-29 at 17.19.17

Screen Shot 2017-03-29 at 21.28.34Screen Shot 2017-03-29 at 21.34.51Screen Shot 2017-03-29 at 21.30.14

The above image is my final logo design. I decided not to have the dots underneath the brand name because I thought that they were too much. I didn’t want to have the ear by itself because I thought that the product may not be recognised as the branding is all new. So I decided to place the name of the brand next to the ear and together they make up the logo. However, I would consider using the word “SYGN” by itself as the brand name.

Product Design

I decided from the beginning that I wanted the product to look like a normal pair of “stylish” glasses. Upon research I discovered a current pair of smart glasses under the name of Vue that already achieved this stylish but simple look. Below are the two types of glasses that they offer.

According to an article from Livestrong.com by Jean Jenkins, “Many deaf people feel that that one of their biggest difficulties the hearing culture that treats them as though they are handicapped, people to be pitied or changed.” I want to create these smart glasses as a piece of assistive technology that is unrecognisable to those who are of the hearing culture. I have realised that if the glasses are made to look like an ordinary pair of glasses then who is going to know that they are in fact assisting the user by helping them to communicate.

On the other hand, it may be argued that by “disguising” the piece of technology as a pair of standard glasses, it’s creating an act of shamefulness. It may suggest that having a hearing-impairment is something that should be hidden. Some may argue that anyone who is deaf should be proud, and by wearing these glasses that “disguise deafness” they are contradicting that idea.

Source: http://www.livestrong.com/article/241936-challenges-that-blind-people-face/

vue

I decided that I wanted a style that was in-between. So I looked further into what kind of styles of glasses are out there. I discovered an article that explains how some glasses are designed according to font faces.

Screen Shot 2017-04-10 at 13.49.52Screen Shot 2017-04-10 at 13.50.58

Out of all of the above, I decided that the style I thought would be best was Helvetica. This style incorporates both the round and the square shape into the frame and therefore I believe it is more likely to suit a wide range of people.

See more at: http://creativity-online.com/work/wk-tokyo-type-glasses-times-new-roman-etc/43140

I created the first mockup design on the programme Illustrator but I wasn’t happy with the turn out because the product didn’t look very realistic.

Glasses Mock Up2

 

So I decided to try and create a 3D model of the smart glasses. To do this I used a programme called Autodesk Fusion 360. I found this task to be difficult at first because I had never used this programme before and I was pretty much teaching myself, I watched a few youtube tutorial videos on how to use the programme and they helped me to start off but as the product developed it became harder to understand how to use the software. Eventually, (after several hours) I managed to produce this as my finished product…

I am aware that the design is not perfect, but I figured that this is a job that can (or would) be perfected by the Industrial Designer. Having said that, I am happy with how the product has turned out and the next thing that I will work on is explaining the features of the product (and their locations).

After researching previous styles of glasses, I decided that I wanted the two temples to be different colours to the main body of the frame. I experimented with primary colours such as red and blue and then moved onto secondary colours like purple and orange, I decided that these block colours were not giving the effect that I wanted. So I thought about using different materials instead of focusing on colours.

In the end, I decided that the main body of the frame would remain as black plastic and the temples would be different metal materials like copper and nickel.

After researching further into the materials used to make eyewear frames, I discovered that Zylonite is most likely going to be the best type of plastic as a material for the main body of the frame as it is super lightweight and extremely cost-effective.

 Image 7 Image Closeup

These are the beginners videos that I watched on YouTube to get an idea of how to use the programme.

Audience

The product that I am proposing will incorporate BSL (British Sign Language) and therefore will only be available to those who communicate through this language.

According to Deafax “approximately 10 million people in the UK have a hearing loss” and “50,000 deaf people in the UK use BSL as their first or preferred language.” This shows that there is a wide audience range for this product.

If the product is successful there will be room for development. The next area to focus on could be America as “there are approximately 28 million hearing-impaired individuals” living in this country and many chose to use ASL as their method of communication.

 

The main audience will be people with hearing-impairments, the product will be available to those of all ages however, the primary target audience will be children (in particular school children) because this product will beneficial in helping children with their education. The secondary audience will be teenagers and young adults, most likely those who are studying at either college or university. Adults may also find the product useful for work purposes.

 

Sources: http://www.livestrong.com/article/241936-challenges-that-blind-people-face/

http://www.deafax.org

Software and Hardware

I began researching further into specific brands of Smart Glasses, I chose to research ORA and Vue because they are both entirely different models, not only do they look completely different, they also have a lot of different features and specifications.

ORA Smart Glasses

ORA-1 are a new pair of smart glasses that have recently been introduced. They are supposedly more advanced than Google Glass and a fraction of the price.

Screen Shot 2017-04-11 at 21.34.22

After reading about the product on the Kickstarter website I discovered that the software that they use for ORA-1 is Android KitKat 4.4.  This mobile operating system, developed by Google, is imperative in order for the glasses to work as smart glasses. It enables the glasses to connect to a mobile device and display data from that device. However, ORA-1 does not work with Android software alone, it needs an Augmented Reality component. This is provided by Wikitude. 

Screen Shot 2017-04-11 at 21.43.38

After researching further into what Wikitude offers, it appears that Wikitude is a company that develops AR programmes for Apps and so it seems that in order for the smart glasses to work in an AR mode, they need to be connected to an application.

Amongst my research on this company, I discovered a price list. These prices are for the cost of the Augmented Reality programming that this company has developed. I think that their prices are realitively expensive and I am pretty certain that the type of programming that I would need for the smart glasses that I am proposing would cost much less as it will not need to incorporate all of the features that this programme does, for example, Wikitude SDK combines 3D tracking technology, image recognition and tracking as well as geo-location AR for apps.

Screen Shot 2017-04-11 at 21.43.12

 

https://www.youtube.com/watch?v=lh7SDw2fYLk

Screen Shot 2017-03-27 at 12.39.32

Screen Shot 2017-03-27 at 12.27.02

 

Below is an image of what is actually inside ORA-1. It is a Patented Retinal Display. This projection technology “allows see-through vision while displaying a super bright virtual image simultaneously. It has a high see-through ratio, wide viewing angle, is light weight, and shatter proof. Furthermore, it is the most cost effective and scalable approach available as all key components are based on molded plastic”. Although this piece of technology appears to be essential in order to create these AR Smart Glasses, it is not something that I would consider for the glasses that I am proposing as I want the glasses to look natural. I would ideally like the lenses to look like normal lenses, so this is something that would have to be taken into consideration when designing the contents of  the product. How is the AR lens going to be made? And where will it come from?

 

Screen Shot 2017-04-11 at 22.18.23

Sources: https://www.kickstarter.com/projects/optinvent-ora1/ora-1-smart-glasses-developer-version?ref=nav_search

https://www.wikitude.com/store/

Vue Smart Glasses

Screen Shot 2017-04-06 at 11.13.29 Screen Shot 2017-04-06 at 11.13.44

 

Smart Glasses

Current Models and Features

Snap Spectacles

  • Simple and stylish.
  • Record 10 seconds of circular video at a time
  • Videos can be shared to social media platforms such as SnapChat or Twitter
  • Relatively cheap in comparison with other smart glasses, selling at $130

Vuzix Blade 3000

  • Arguably the nicest looking and most comfortable
  • Offer a display projected onto the right lens to overlay emails and even YouTube videos in the world around you
  • Voice assistant
  • Touchpad controls
  • 8 hours battery life
  • Camera to capture stills and 1080p video
  • Bluetooth
  • Wifi
  • Costs less than $1000

ODG R7/R8 and R9

  • 720p lens
  • 4MP camera
  • Voice recognition
  • Wifi
  • Bluetooth
  • R7 costs $2750, R8 costs $1000 and R9 costs $1800

Vue

  • Product of a $2 million Kickstarter campaign
  • Doesn’t use Augmented Reality
  • Uses bone conduction tech so you can use the glasses instead of headphones
  • Touch interface to control music and calls
  • Enables you to handle notifications and activity tracking
  • Priced at $189

Cast AR

  • Camera that scans your surroundings
  • Projects images through two micro-projectors installed on the top of the frames
  • Priced between $290 and $765

Moverio BT-200

  • AR headset enables users to watch HD and 3D contents
  • Front facing camera
  • Motion sensor
  • Built-in Dolby Digital Plus for sound
  • GPS
  • Compass
  • Microphone
  • Projectors
  • Projects images at 960×540 resolution to transparent glasses
  • Not wireless – needs to be connected to an Android device at all times – that’s where it’s computing power comes from
  • Lasts up to 6 hours
  • Specifications such as 1.2GHz dual core processor and 1GB RAM
  • Priced at $699

Meta

  • Focuses on what google glass does not do
  • Overlays AR on top of your reality
  • Allows user to manipulate objects
  • Motion tracking
  • 3D HD display
  • 3D surround sound
  • Camera and quality lenses
  • Costs vary from $667 – $3650

Researching current models and their features was very useful when it came to thinking about and putting together my own product. I had already thought about the basic principles of a pair of smart glasses but there were some aspects that I forgot about, such as the importance of a long battery life, the amount and/or position of the micro projectors. I decided after researching the existing products out there that I would sketch a rough drawing of the glasses and begin to think about the features that I wanted to include.

Also, from my research I learnt about the audiences to which the smart glasses are aimed at, I realised that the majority of smart glasses are designed for cyclists or for gym purposes. Very few are designed for greater needs, such as helping a person who is less-abled. However, I did come across one pair of glasses that could be adapted to suit the needs of someone with a hearing impairment, these were called GlassUP.

 Draft Mock Up Drawing of Glasses and Features

fullsizeoutput_222

Deaf Culture and Sign Language

The Issues

As a person who can hear and has never experienced deafness, it is difficult to understand what it is like living with a hearing-impairment. I decided to do some research to try and uncover what difficulties people with hearing difficulties have to face.

Screen Shot 2017-04-16 at 14.42.07

The topics raised in this discussion are topics that would need to be covered through the development of the project. The smart-glasses could potentially solve the issue of having no or unqualified sign language interpreters, because the project will require a highly qualified professional interpreter to work alongside software engineers in order to manufacture the right piece of technology that can equip this proposal. Also, the product could potentially resolve the issues that occur from lip reading as it offers an alternative solution.

Source: https://www.quora.com/What-are-some-problems-that-Deaf-people-face-on-a-daily-basis

Sign Language

I decided to research the different forms of sign language available in order to determine which method of sign language would be appropriate for the project. I found a blog from The Disability Support Services program of Gulf Coast State College which describes that there are many different types of sign language and each one differentiates slightly from the next.

After reading the first paragraph, I already began to understand a little bit more about the complexity of sign language. “Sign language uses visually transmitted sign patterns to convey meaning by simultaneously combining shape, orientation and movement of the hands, arms or body, and facial expressions. This is important as a sign not only conveys a word but also conveys a tone, both contextually and acoustically.” It was then explained that the “emphasis on the word, whether it be delivered by using a higher or lower pitch, the volume, and even tones such as sarcasm are all being conveyed simultaneously by the interpreter with use of the sign, body movement, and facial expression.”

From this reading I understood that sign language was not straight forward. Not only are there different variations, there are also different portrayals. Some variations use hand, face and body movements to make up the language, and others use the hand, face, body and head. I began to realise that sign is a very complex language. I realised as well that the tone of ones voice can be conveyed through sign language as well as the words spoken. Also, I understood that there were different systems of sign language that were often preferred in some schools.

British Sign Language (BSL). Though the United Kingdon and the United States both share English as a spoken and written language, British Sign and American Sign languages are quite different from one another and are not mutually intelligible. British Sign Language (BSL) makes use of space and involves movements of the hands, body, face, and head.

American Sign Language (ASL). ASL is a complete, complex language that employs signs made with the hands and other movements, including facial expressions and postures of the body. It is the first language of many deaf in North America, and is considered the dominant sign language of the Deaf community in the United States, in the English-speaking parts of Canada, and in parts of Mexico. ASL is said to be the fourth most commonly used language in the U.S.

Signed English (SE). This system is preferred among primary and higher education institutions as it has one sign to represent each word in the English language. It is intended to be used to help with reading and writing, and has important signs to teach grammar.

Sign Supported English (SSE). This is the preferred method of hearing people to communicate with the deaf. SSE uses the same sign as British Sign Language (BSL), but unlike Signed English (SE), you do not have to sign every word. SSE also doesn’t have its own grammar system like BSL, enabling hearing people to use the sign system without having to learn a whole new grammatical structure. SSE can be picked up fairly quickly which expedites communication.

International Sign (IS). You may not come in contact with this system much here on our campus — yet, that is; we are growing — but IS is an international auxillary language used at international meetings, such as the World Federation of the Deaf (WFD) Congress and events such as the Deaflympics.

Padget Gorman Signed Speech. This signing system is used with speech to help those with langauge difficulties. There are 37 basic signs which when combined can make over 4,000 more complex ones.

Pidgin Signed English (PSE). Another characterstic of language, like vocal language, sign language has a very crude signing system, in which elements of BSL and spoken English are combined to allow communication between hearing people and deaf people who only know the strict confines of sign language. This system is not recommended (the persnickety of proper grammar should cover both ears and eyes for this system).

Finger Spelling. This signing system is generally used alongside sign language. It is used to spell out names, places, and anything else where there is not usually a sign for. Many times, new words take longer to be adapted into a singular sign, therefore they must be spelled out.

 

After considering the potential sign languages available, I believe that either Signed English (SE) or Sign Supported English (SSE) would be the best option, this is because the aim of the product is to translate spoken language word for word and some sign systems may make it difficult for this to accomplished.

https://dssofgcsc.wordpress.com/for-professors/types-of-sign-language/

Further research confirmed that “Signed English is a sign language dialect which matches each spoken word of English. It is mostly used for language development, allowing a teacher to reinforce the spoken word with its equivalent sign. Children are able to communicate in Signed English at a much earlier age than the spoken language; transition into to spoken English is made easier as they already possess the signed vocabulary.”

http://sign.com.au

The sign.com.au website contains a complete dictionary of almost 3,000 words which means that it can translate words or phrases into Signed English.

Screen Shot 2017-04-17 at 14.43.33