NOTE: Images can be expanded by clicking on them

———————————————————————————————————————————————

Welcome to the Project 3 page. This page contains all information on the blind sensor aid device/detector device that I am working on.

Project aim:  to design a device which blind people can use as an aid to detect and visualise items and their features.

Project Objectives

I will plan these out on the Learning outcome plan. All changes and updates will be on the Level 5 LO. plan page in the latest version of the plan.

• Identify the needs and wants that blind people may want from a device that helps them visualise items and their features.
• Take apart a Talking colour detector and analyse it, to see how it functions and performs.
• Make and identify the risks using risk assessment
• Using a video from YouTube and an Arduino with components, attempt to make my own simple colour sensing device similar to the YouTube video.
• Analyse and test colour sensor to evaluate how it works and what impacts the values.

Context

Across the country, there are many blind persons who can partially see or not able to see at all. There are many different types of blindness such as central blindness or peripheral blindness. Some blind people like to know more information about their surrounding area and because of this, some use a talking colour detector. This detects the colour of an object and communicates this to the blind user. Blind people may want other aspects to be relayed to them, giving them more information on the world around them. Also some may want other features to more benefit them.

——————————————————————————————————————————————–

Requirements:

Performance requirements:

• The system needs to at least be able to detect 1 feature.
• The system must function correctly with more than 80% accuracy in its detection.
• The system must not weigh more than 600g.
• The user must be capable of hearing the audio clearly.
• The system must be comfortable to hold.
• The system must be able to detect distance, at least 50mm as a minimum maximum.
• The system must hold the sensors and the components secure to prevent damage from drops.
• The system must have a unique shape so it is easily identifiable without the need of sight.
• The system must speak its results to the user.
• The system must be able to be turned off or stopped.

Manufacture requirements:

• The system should be easy to assembly and disassembly for maintenance
• The system should not have any shape edges and should be comfortable for holding
• The casing for the system should be a material that can support and prevent damage to the components

Operation requirements:

• Components in the system must be easily accessible for maintenance
• The system should be portable via battery power or have changeable capability
• The system must be able to detect brightness or have set brightness’s i.e. day and night settings
• The user needs to be able to operate the device with ease.
• The user needs to be able to operate the device without the use of sight.
• The user has to be able to gain information from the device without the use of sight.

Acceptance standards:

• The system’s electronics should be tested before being integrated into the system

• The system must follow BS EN IEC 62485-1:2018 standard for safety requirements for specifying, installing, commissioning or using batteries if a battery is implemented

Disposal requirements:

• The system should be handed to a specialist waste sight because of the electronic components inside
• If the system is faulty then it should be returned to the manufacture for proper disposal

———————————————————————————————————————————————–

For this project I am going to start by taking apart a talking colour detector as I already have a couple with me: I currently have two types, one older model and two versions on a newer model. I will also do a risk assessment and assess the dangers and risks. This will start on the 6th of November 2021.

Disassembly of Talking colour detector:

For this task I decided to start by disassembling one of the newer versions of the talking colour detectors. For this I used screwdrivers of different head sizes as there were four screws that were needed to be removed to open the casing.

How does the device work:

The device works by first sensing whether there is an object in front of the device. If there is, this triggers the colour sensor to start. This components detects the amount of each colour in the object it is sensing. Different colours have different red, green and blue values. Depending on these values, a signal will be sent to the loudspeaker and it will speak the corresponding colour.

The object sensor prevents the system from getting confused and only senses colour if an object is present. Not having an object sensor in the system would cause it to constantly speak random colours or make beeping sounds because it cannot pick up a colour.

The switch is slightly different. Unlike normal switches that only have on and off, this one has a middle part that allows for calibration.

Other components of the device:

Potentials for my own version:

• There is no volume control so it is constantly at a high volume which can be painful if held too close. Volume control may be a good feature to have so it is more suitable for the user (potentially helping with hearing problems too).
• A calibration button is currently needed. maybe I could design mine so it is already calibrated and this isn’t required.
• Like this version, my version could have a switch which helps calibrate if needed. This could help blind user as they wouldn’t have to memories the location of multiple buttons.
• A sensor focuser may be required to allow for more accurate readings of colours

Information on chips used in talking colour detector:

LM358N chip datasheet: https://www.ti.com/

24LC256 Chip datasheet: http://ww1.microchip.com/downloads/en/devicedoc/21203m.pdf

What I have learned

I have learnt how a talking colour detector functions and how different components work to ensure accuracy within the device and how they can work with the other components to function as intended.

Learning outcomes involved

DE502 LO1: For this task I selected and chose to use screwdrivers of different sizes. This was so that I could open the device. Also there was no need for any other tools as there was only screws that was holding it together.

DE503 LO3: For this task I used IT to write up my findings and I evaluated the system as well.

———————————————————————————————————————————————–

Risk Assessment for making and testing

Risk issues

• There is a risk of failure to complete the project in the designated time frame.
• There is a risk of illness preventing progress in the project
• There is a risk information cannot be distributed due to illness, delays or others (questionnaire distribution)
• There is a risk of electrocution from damaged or high voltage components in the project
• There is a financial risk. Not having enough budget to complete or have certain parts for the project.
• There is a risk people could be offended or have there details be exposed to others that they may not want.

Managing risks

Risk of failure: this can be managed by planning out and mapping each of the tasks within the project.

Risk of illness: this can be managed by keeping on top of health and safety, staying clean and ensuring correct use of precautions.

Risk of information distribution: this can be managed by using electronical alternatives such as email or text.

Risk of electricity/electrocution: this can be managed by replacing/not using damaged components. PPE could be worn to prevent electrocution too.

Financial Risks: This can be managed by ensuring you sick to a budget and find alternatives to also keep within it. Analysis using cost models and pareto histograms among others to reduce costs and ensure no financial problems.

Ethical and personal details risks: This can be managed by ensuring anonymousness, consent and letting the people involved know that they do not have to participate in parts of the project if they which not to. make it clear that this is optional and they are not required to do it.

———————————————————————————————————————————————–

Ethics

For this project I wrote up a BREAM ethics application. In this I highlight any ethical issues that there may be in the project as well as tasks that I will be taking and ethical issues that may occur. Here is a link to my BREAM application:

Click Here: BREAM for Blind sensor aid project

———————————————————————————————————————————————–

Concepts

Concept 1:

Here is a photo of the first concept a came up with for the blind sensor aid device:

The idea for this one was that it would function a bit like a torch. The sensors would be or the end with a cone like filter which could help focus the sensor. The main components of the system such as the microcontroller would be in the central box structure. This would also house the buttons on its side, a screen on the top and speakers either on the other side or on the top with the screen. Components may also be in the handle to save space. This concept would use batteries which could also be held in the handle. The handle would allow for ease of use and more comfortability when using it.

Concept 2:

Concept 2 is designed to be flat and similar to a smartphone. Its extra thin design would allow it to be easily carried and held. The speaker, buttons and screen would all be on the top surface and the sensors would be on the bottom. With this design there would be no limits to orientation or distance when sensing.

Concept 3:

This is a box design which, instead of holding it or aiming it at an object, it is placed on the object. This version would have a fixed distance and because of the sensors being far instead the box, the brightness would also be fixed. This would eliminate these factors and would allow for more accuracy, however, it would have less places in which it could be used. The box like shape would also allow for more stability and secureness of the components. The screen, speaker and buttons would also be on the sides and top of the concept.

Concept 4:

Concept 4 works with the main body holding all the components being attached to the user. The sensors would be attached to the main body via a wire or cable. The sensors would be aimed at or placed on objects by the user. This design could work on the belts or clothes of users however it would not be able to have a screen because of this and the speakers would need to be louder. Not having a screen is not 100% necessary though because the target audience are blind people but if they are partially sighted they may want a screen.

Improved concepts

Once I had drawn up the concepts I went back and improved them, fine lining the outlines, adding shading, labels and giving the concepts names. By doing all of these, the concepts look much better and are more easily understood especially with the hand and showing how the device is to be used. This provides readers with more information than just a simple picture. Below are the improved concepts:

Final improved Concepts

Once I had finished the improved concepts I then went on to add some colour to help the concepts stand out and pop.

Here are the concepts that I improved. I also made a final improved concept for the last concept since I had not draw one. For these concepts I tried a mix of background colour or having the concept itself coloured.

———————————————————————————————————————————————–

Decision matrix for concepts

Here is the decision matrix that I made for the concept I had drawn up:

The winner was the talking precision cube concept scoring the highest.

———————————————————————————————————————————————–

Questionnaire

For this project I needed to find out what the device that I am building should detect. To discover this, I decided to make a questionnaire which would allow me to understand the needs and wants of potential users of the device. Here is a photo of the questions of my initial questionnaire.

As you can see, these questions are basic and not specific, the layout is bland and overall this would not provide good information that I could evaluate and use in the future.  Also the questionnaire had nothing on it about consent or how the information was going to be used. After discussing with other design engineers and lecturers, I went on and researched into how to improve my questionnaire. I researched into types of questions, layouts and how to section it.

Here is a photo of the improved questionnaire.

NOTE: For the image and the below download I have removed my email to protect my privacy but for the actually questionnaires sent, my email was included.

Here is a download link to the full version of the questionnaire: Click Here

Improvements made:

I started by adding a border around the questionnaire and page numbers to fill in the white space. This helped made the questionnaire look more professional and presentable. I implemented more questions and sectioned them so they were all about the same parts. For example, questions in section 1 are about the current devices. I also made the sections have the same style of question throughout (section 1 was all 1-5 questions, section 2 was mostly ranking). A couple paragraphs were added talking about the questionnaire’s anonymousness, the consent of the questionnaire and it doesn’t have to be completed or answered if they wish not to. I have provided my email and a general overview about the questionnaire.

Ethical issues

An ethical issue that there is with the questionnaire is that the blind or partially sighted may not be the ones writing down their answers. In many cases, it will be a carer that writes the answers down for them. This is a problem as this could lead to the carer controlling the answer and potentially preventing the people from answering how they want to. Another issue is that the carer, if they did not want to answer for the blind they could not write anything down and tell the blind person that they have written their answer (the blind participant would not know this has happened).

——————————————————————————————————————————————————————–

Questionnaire results

NOTE: To protect the privacy and the anonymousness of the participant, the results of the questionnaire cannot be shown on this webpage.

From these questionnaires it seems as though some blind users are quite happy with the current devices out there but systems such as a low battery system and a volume dial may be a way of improving these. In terms of other features that the device could detect, There was a suggestion that the device could detect and identify logos. I believe this could be a very good feature to implement because many blind people may want to buy specific branded items my would not know unless they asked. This would especially be good for clothing and boxed items.

Goal:

The goal now is to construct a functioning colour detector and once this has been completed work towards adding features such as the volume adjustment and a battery measurement system. Once this is complemented then the identification of objects would be the next set of features. The features should be reviewed to see which ones would be the best to integrate taking into account factors such as cost, time and availability.

——————————————————————————————————————————————————————–

Arduino code and build

To get the colour detector working I decided to use Arduino. For this I would start by displaying a number value, which the light sensor obtains, on a LCD screen. I started by doing some research into how to build a colour detector using Arduino. This provided me with some code and a circuit diagram that I could follow. Below is the circuit diagram that I used to build my colour detecting system.

https://create.arduino.cc/projecthub/mjrobot/arduino-color-detection-57e4ce

Here is a photo of the build I constructed. It includes an Arduino UNO, a LCD screen, a potentiometer, wires, a button and the TCS3200 colour sensor. The potentiometer is to control the brightness of the LCD screen. The TCS3200 colour sensor is what detects the colour. It does this through an array of 64 photodiodes which are covered in four filters and detects different light incidents. sixteen photodiodes filtered so they can only detect red, sixteen for blue and sixteen for green. the remaining sixteen have a clear filter. The LCD screen is to display the results detected by the colour sensor and the button is to start the LCD screen and the detection.

https://www.elecrow.com/wiki/index.php?title=TCS3200_Colour_Sensor_Module#:~:text=TCS3200%20chip%20is%20designed%20to,red%20in%20the%20incident%20light.

For my build the circuit would be slightly different, involving a potentiometer and an Arduino UNO instead of the Arduino Nano on the diagram. To begin I started by taking the code provided and entered this into the Arduino but quickly found that it didn’t function correctly with my build. However, I was able to use parts of the code provided and some help code on Arduino.cc to get the build functioning. Below is the original code that I managed to produce in order to display on the LCD screen the colour values that the colour sensor was detecting. In the end this code did not function as intended.

Here is the current code that displays on the LCD screen the colour values that the colour sensor is detecting:

Here are the links to the code and other resources that I used to help making my own:

Code, pin out and circuit diagram Fritzing build:

https://create.arduino.cc/projecthub/mjrobot/arduino-color-detection-57e4ce

https://create.arduino.cc/projecthub/SurtrTech/color-detection-using-tcs3200-230-84a663

basic Arduino code for components:

https://docs.arduino.cc/learn/electronics/lcd-displays

https://www.arduino.cc/en/Tutorial/BuiltInExamples/Button

 

Below is a photo of what this code displays on the LCD screen:

——————————————————————————————————————————————————————–

Technical report on RGB values and multiple regression

Here is my technical report on the RGB values and multiple regression that was used to help constructed a functioning and accurate colour detector system. This technical report includes RGB graphs, repeat value tests, individual value tests, multiple regression, checking system and a flow chart. This report also includes the Arduino build and the Arduino code that was built and used.

Click here to view and download: RGB value and multiple regression technical report

The below sections have more information and downloads to some of the work in the technical report.

——————————————————————————————————————————————————————–

RGB colour-distance graphs for TCS3200 colour sensor

The aim of making these graphs was to help determine a suitable distance that the colour sensor should sense colour from. It was also to give it its own colour range as the values the component produces doesn’t match standard colour RGB values. Waveshare, the company that produces the component, recommended 10mm from the colour its detecting and have a voltage between 2.7v-5.5v. For the making of a basic colour detector I decided to go for the 3.3v on the Arduino as this was roughly in the centre of the range and it wasn’t too bright.

here is a link to Waveshare’s specifications page for the colour sensor:

https://www.waveshare.com/color-sensor.htm

The TCS3200 colour sensor components works slightly different to the standard RGB colour values. The normal values range from 0 to 255 but this components does not. It rarely goes above 100.  If I were to code it so that it functions with standard colour values, it would not work and would read the incorrect colours. For example, the standard values for black are close to zero however for the colour sensor they are close to 40/50. If I were to convert it (255- colour values) this would as that black is about 200 for all its values which is incorrect.

Graphs produced: RGB colour distance graphs can be seen in the technical report that I have made.

Here is a link to the full excel spreadsheet with all results:

RGB colour graphs

———————————————————————————————————————————————–

Repeated colour Values

To find out how accurate the colour sensors are, I decided to do some repeatability tests. This would provide me with multiple results and I would be able to make graphs comparing all the red values, green values and blue values with one another. To ensure fairness between the data collecting, the device was placed in the same location, the lighting was the same and the same piece of paper was used.

Graphs: Some graphs can be viewed in the technical report or in the excel downloads below.

Here is a download link to the graphs and the values obtained from the device: Repeat RGB values v3

Here is a download link of a previous version:

Repeat RGB values v2

Repeat RGB values

——————————————————————————————————————————————————————–

Multiple regression

For the Arduino build and code to detect and display the correct colour, it needs to take into account brightness, distances and the raw red, green and blue values. the best way to do this is through multiple regression where I can obtain an equation which can be used in the code. For this task I would need to collect all the data needed and display them in a table. For the raw values of the colour I took the average of the repeat values that I obtained. I started by making two tables of data, one which included brightness and one which did not. Here is a photo of the data for the without brightness table:

So that the colours matched their data sets and some they could have a value for Y, I gave them corresponding numbers. To perform the multiple regression of this table I used the link below to help learn how to perform it in excel. Previously, I had used a website, also linked below, in which you place your data into the set table and it would produce an equation. The problem with this was that there was a limited amount of data sets you could have. not only that but the excel version has no limits and will also tell you whether a certain data set was necessary or not.

How to perform multiple regression website:

https://www.statology.org/multiple-linear-regression-excel/

multiple regression tool:

https://www.socscistatistics.com/tests/multipleregression/default.aspx

Here is a photo of the information that was generated by performing multiple regression through excel:

From this data I used the “How to perform multiple regression” link to produce an equation. This is done using the coefficients and the intercept. The intercept is when all values are zero (zero distance and zero colour values). The standard error is how far from the regression linear on average they are, P-value can help determine how important the data set is and can tell you how accurate and significant it is. If the P-value is less than 0.05 then the variable is very significant.

To test the equation and the multiple regression, I decided to perform some tests by which I would place in expected values for an expected colour and see what colour the multiple regression would predict. Here is a photo of a table I made to do this for the data without brightness included:

I then decided to go a bit further to see exactly here the regression was guessing incorrectly and correctly. Here is a photo of this:

This data shows that the multiple regression is able to guess black, blue, green and red quite well with all of them guessing 3 or more correctly. some values are slightly off, with it guessing the colour that has a value below it. For example one of the guesses should have resulted in black but it guessed purple. The value it produced though was only 0.12 away from it correctly guessing black. If a tolerance is built into the system, then it would allow for this type of event. Another example is the green 40mm and 50mm test, it was only over by 0.03 and 0.07.

in total it guessed 24 out of 60 different value sets which means it has currently guessed 40% successfully. This will alter with more data and with brightness included.

Full data sheet with brightness and distances 10-50

Once I had collected all the data, all night and day values for each distance for all colours, I put it all into a spreadsheet table similar to the first set of data. Here are some photos of this table:

Here is a photo of the information generated using this table and by performing multiple regression. I have also include a table saying what value Y should be for each colour:

Once I had obtained an equation I, much like the first set of data, went to test how accurate it is. I noticed that it wasn’t picking up the same colours as the first set of data. For example, when you entire values in that should output a purple, it would output a yellow result. Colours that were less affected by this were red, green, blue and black. I think this is because of how the sensor works, having red, green, blue and a clear filter covering sixteen photodiodes each. This could be why it isn’t picking up the other colours as well. The success percent of this equation was 35%. This meant it was guessing the majority of inputted values, incorrectly.

Checking and patterns

To increase the amount that the equation successfully guesses correctly I either had to obtain more data that covered more of a range or I could implement a checking system into Arduino. I chose the second option because I had noticed there was some patterns with the numbers. For example if I set the G value to zero, all white values ended up being within a given range and no other colour would be within it.

Here is a download link to the entire excel spreadsheet, this will be updated until complete:

Click Here: multiple regression sheet v5

older versions:

Click Here: multiple regression sheet v4

Click Here: multiple regression sheet v3

Click Here: multiple regression sheet v2

Click Here: multiple regression sheet v1

Checking percentage increase

Whilst finding the patterns for the checking system, I mapped out the percentage increase. This was important to do as it would allow me do see whether changing and applying the checks would increase or decrease the accuracy of the system. If I had not done this then I may have decreased the the amount of successful guess and would have only know write at the end. Below is a graph of the percentage increase with each pattern and check that I implemented:

Before the checks the percentage of successful predictions was 38.25% and after the checks that went all the way up to 86.25%

What I have learnt

I have learnt how to perform multiple regression and how to interpret the data to produce an equation.

——————————————————————————————————————————————————————–

Flow chart for checking system

to integrate the checking system into the code for the Arduino build I made a flow chart. This was so that I could write the code in the correct order and with ease. Below is a photo and a download link to the flow chart. I have put sections of the flow chart in different boxes. This is to show what the initial Y value is and what colour it initially predicts.

Download link Click Here: flow chart for multiple regression

——————————————————————————————————————————————————————–

Displaying written colour

To start this task I needed to alter the basic colour detector code that I had assembled and change it so that instead of displaying the values of each colour, it would display the word for the colours. I started by using the RGB values that I had collected and added some ranges and IF statements so that when the colour sensor picked up a value (this is now done in the background and does not display on the screen) it would check what range it was in and display the corresponding colour. For this I have started by only using the red values that are picked up, using a fixed distance, only using the primary colours and a fixed brightness.

Here is a photo of the code that I created:

Here is a photo of the display:

Next I decided to alter the code again but this time include the values for red, green and blue instead of just red. The distance is still fixed and so is the brightness. Here is a photo of the code for this part:

Potential problem: when doing the RGB value tests, green at 10mm distance had a constant value of 10. Because of this, in the coding I have put that Greendata has to equal 10 instead of a range. This can be problematic if the sensor picks up the green as a different value, ultimately leading to the wrong colour being displayed or a colour not being displayed at all. Having ranges would be slightly better for this as a slight difference in the value could be allowed.

——————————————————————————————————————————————————————–

Full Arduino build and code for functioning colour detector

For the Fully functioning colour detector Arduino build, I decided to make my own circuit schematic using Fritzing as I have implemented many components that where not in the original circuit schematic I was using. For the colour sensor, I had to use a different model because the program didn’t have the one that I used in the build. Here is a link to the download I used for the component: https://forum.fritzing.org/t/color-sensor-tcs230-color-sensor/14922

On the schematic there is a wire that goes to the colour sensor but doesn’t connect, on the model I used this would go the the LED pin. Here is a photo of the fully functioning colour detector circuit schematic:

Here is another photo of the circuit but this version is a circuit diagram. Because the colour sensor was imported the schematic does not have any pin points and I was unable to add any. If I had added pin points or detached the wire, then it would impact and remove parts from the circuit schematic. Because of this there is only one pin point and the schematic should used to see the correct pins of the colour sensor.

Below is an image of the full Arduino build including the ultrasound module:

Once I had built in all the components to the Arduino build I started to alter the Arduino code that I had so that it included the multiple regression and the checking system. To do this I used the flow chart that I made. Below is the full code that I produced and used to operate the system:

NOTE: Please go to Arduino full build page to copy code for own build

 

Here is a short video of the colour detector system functioning:

The colour detector still picks up incorrect colours occasionally but it is much less than what it use to. It is now more capable of sensing the correct colour. I have noticed that sometimes you need to jolt or change the item you are detecting in order for the system to start. It can also be started by pressing the button.

——————————————————————————————————————————————————————–

Current suitability of device

Currently the device is not suitable for the blind user. Firstly, the wiring is exposed and not secure. This means that one knock of the wires can cause the device not to work. Unless you know how to use the components, you wont be able to rebuild the device. It would be especially difficult for the blind. Another reason it currently isn’t suitable is that the device can only be turned on by connecting a USB. This has very limiting range which can mean the user could not really take it anywhere. A battery system was planned so that the device could be more portable but due to time this sub-system has not currently been implemented. The system also displays the results of the detection on a screen. This has been helpful for testing, however, it is not suitable for the users as they are blind. A loudspeaker system needs to be added to make it suitable.

The current device only detects colour. It only detects primary, secondary and shades but nothing in between, for example, light blue or dark grey. Furthermore the device hasn’t detected anything other than colour and some of the ideas were for the device to detect more than just colour. So far the device is not suitable for users  but if many other systems are implemented then the device could be very suitable.

Systems that would make the device suitable:

• loudspeaker system
• battery system
• button/switch system
• systems for detecting other features

——————————————————————————————————————————————————————–

Proposal of Project

I propose that I take this project further in the future. So far this project has provided a colour sensing device which can be used to aid blind people. This project can be used to further assist blind people more by building different features into the system which could detect other items, logos, types of objects etc. I would then propose that this device be patented and commercialized for the blind to use. An idea that cropped up when speaking to others about this project was using barcodes to identify specific items. It could result into this and use this going forward (maybe own barcodes that can be stuck on items).

Patent:

Patents can be important as they can prevent others from producing and manufacturing your product. It can be used to protect your work or invention. Likewise, they can also prevent you from producing your product if you are making a product similar to someone else’s. It is important that we have a look around to see what patents other people have because it will lower the chance of you infringing on other peoples products and inventions. If we do not do this and we produce a product to sell, we could be infringing on other peoples work and this can result in law suits and fines. Infringing could cause harm to the patent owners business and work such as loss of profits.

To register your design this link can be used:

https://www.gov.uk/register-a-design

Here is a link to some rights that will protect your design even if you do not register your design:

https://www.gov.uk/unregistered-designs

——————————————————————————————————————————————————————–

Quality issues

• One quality issue is the components of the build. For example, an Arduino Uno board may say 3.3v but some may actually output something like 3.35v. This may cause the LEDs on the colour sensor to shine brighter or dimmer which can ultimately lead to altered values and a decrease in the accuracy of the designed device.
• Another quality issues is that the accuracy of the device can be altered if the code isn’t correct or used in all devices.

Quality planning

Components should be tested before constructing the device. This prevents faulty components from entering the devices.

Components should also be tested once the devices has been fully built. This will check as see if any of the components have been damaged during construction and it will identify if a components has become faulty due to other components.

Features should be tested once the code has been implemented into the device. This will ensure quality and check everything is working as intended.

Quality assurance (preventing defects)

Components should be tested before being built into the system. This is so that they function correctly and at the correct properties. This will ensure accuracy is constant throughout multiple builds.

The same code should be used in all devices made. This will prevent differences in the devices and will ensure quality.

Machinery used to construct the device should be the same. This will ensure the devices are all built the same and with the same components. Components should be brought from the same suppliers as buying from a different one may cause differences and inaccuracies.

Quality control (meeting standards)

If components are faulty and don’t function as intended, they should be removed from the system and replaced with a proper functioning components. This will ensure safety with the device, longevity and increased quality.

Quality improvement

The quality can be improved by more research into the features and components. This would allow for more accuracy in the device. Once research has been done, the a new updated code can be made. This would then be implemented into a new batch of the devices. An integrated feature on older devices could be having the Arduino port available. This would allow for the older devices to be updated with the new code.

——————————————————————————————————————————————————————–

Commercial, social and economical

Commercial:

A way this project could be commercialized is by setting up a website that people can go to. This webpage would contain a pack which people could download and use as a guide to construct there very own. Because of some blind people cannot see at all and some are partially sighted, this would have to be built by family members, making this option not the most suitable. Not only that but this site would contain links to websites which they can buy the components and download the programs. This would require the people to understand how to do this and may make it very tricky for some.

another way to commercialise to be to speak to suppliers of the components about this product and potentially strike a deal where either they may invest in it or would help build. This may also potentially result in companies producing specialised components for a better version of the product.

Communicating with charities is another way. This could work by the charity communicating the product to the blind people as an aid. People who are interested could request for one which would then be constructed on demand. The money for the components may be provided from donations.

Social:

This project and product could be advertised using social media. This may not be the best approach for this project as the people this device is for are the blind. This means that only some blind or partially sighted people would be able to see the advert. Also for those who cannot see the advert may be told by the relatives or the carers of the blind person. Another way the product could be advertised is through working with charities. There are many charities that provide aid to blind people and when they are in communication with the blind, they could mention it as an optional aid. This way the blind people would gain greater knowledge of the device along with having a potential way it could be supplied to them.

Economical:

For this product, the money necessary to build the device could come from donations from charities, own investment or investment from companies that believe the product is profitable. Working with charities would be good as the product could be made on demand with the components being brought using donations. This would mean that the money is going into the product which would then be going to aid the blind user. Sometimes when money is donated to a charity it does not always aid the cause and by doing the product this way it will ensure the aid of the cause.

——————————————————————————————————————————————————————–

Codes of conduct

• Equality – people should be treated equality and with respect. disrespecting others in not okay and can harm your reputation and companies that you work for. Whether people are disable, of different gender, race and others, everyone should be given the same treatment and opportunities
• Discrimination – people should be treated with kindness and should not be discriminated against. This is not okay in the workplace
• The work that is completed should not harm anyone of any culture, religion or race. Work should be suitable for the majority of people.
• Truthful – You should be truthful to the people that you are work with. Not being honest can drastically harm your reputation and the people around you.
• Behaviour – We must behave correctly. Violence, insulting and harming others is not okay and can impact yourself and the people around you.

——————————————————————————————————————————————————————–

Standards and legislations

Below are some standards and legislations that should be followed when working on the product. It is important that we take these into consideration because it could cause harm to others if we didn’t and we could be breaking the law and reals.

• BS EN IEC 62435-8:2020 – Electronic components. Long-term storage of electronic semiconductor devices. Passive electronic devices

• The Waste Electrical or Electronic Equipment (WEEE) Directive – this requires countries to maximise separate collection and environmentally friendly processing of these items.

• The Electrical Appliances (Colour Code) (Amendment) Regulations 1977 – this requires a label be attached to the main wire explaining the different coloured wires in the appliance.

• Provision and Use of Work Equipment Regulations 1998 (PUWER)
• Portable appliance test (PAT) – electrical equipment that is portable has to be tested before being used, distributed or sold.

——————————————————————————————————————————————————————–

Reference sheet

Here are some reference links that I used for this blind sensor project

Arduino code and build references

Fritzing build:

https://create.arduino.cc/projecthub/mjrobot/arduino-color-detection-57e4ce

Colour sensor code and pin out:

https://create.arduino.cc/projecthub/SurtrTech/color-detection-using-tcs3200-230-84a663

basic Arduino code for components:

https://docs.arduino.cc/learn/electronics/lcd-displays

https://www.arduino.cc/en/Tutorial/BuiltInExamples/Button

TCS3200 colour sensor information:

https://www.waveshare.com/color-sensor.htm

how it works: https://www.elecrow.com/wiki/index.php?title=TCS3200_Colour_Sensor_Module#:~:text=TCS3200%20chip%20is%20designed%20to,red%20in%20the%20incident%20light.

Colour chart:

https://www.rapidtables.com/web/color/RGB_Color.html