Progress Log

Week of 4/19/2021

This week, we worked on finishing the work on all three devices by attaching everything inside and testing all the app functions to make sure they worked like a complete product. This included fixing problems with the startup code and fixing any bugs by running through each app individually and all at once(like the customer would) to make sure there weren't any problems during the presentation. Lastly, we made sure to add our testing data of each of the apps to the report.

Week of 4/12/2021

This week, we worked on testing the braille mechanism on the assistant box with the new converter and the wiring scheme for powering the agents and were successful in avoiding any low voltage warnings on the pi. We also finished soldering all the wires for the magnets in the assistant box and tested the input voltage level to the pi from the converter with a USB tester to make sure it could supply enough voltage and current while powering the magnets. Lastly we did a power consumption test with a multimeter by running the braille display through the alphabet and made sockets for the converter and raspberry pi/UPS so it is possible to unscrew them to do maintenance while also adding magnet latches on either side of the box so we can open or close the lid.

Week of 4/5/2021

This week, we worked on testing the braille mechanism with the raspberry pi but ran into problems when we got a low voltage warning when we had 5 magnets on at the same time. Although it works, we decided to power the magnets from the power source directly instead of through the pi to avoid drawing too much current. To do this, we tested powering the magnets through a new wiring scheme and are using a new converter that can power the raspberry pi and the magnets separately, also the wiring for the necklace is done.

The code for the heart rate oximeter and thermophile readings is done. The face meeting app is now in the main program where there is a combined function to detect faces in camera view, an OCR function was added to the main script. The startup sequence was updated with language settings, braille or audio mode and set the input/PIN code. Lastly the data transfer in the device was changed from using text files to executing scripts with data as arguments whenever possible which increased efficiency by 84 percent.

Week of 3/29/2021

This week, we continued to work on the updated design of our main assistant box by adding ports for the ups charger and holes for the power switches, made a mesh cover on the side to dissipate heat and allow sound waves to enter the box for our USB mic and made a spot for our magnet latch to be attached so we can open and close the box. Furthermore, we added mouse ears on the sides of the box to prevent the 3d warping we got last time we printed the box..

For the wiring on the dial glove, we were able to stop the wires from popping out of the raspberry pi by using a more flexible electrical wire and have finished the soldering that to the inputs on the Darlington driver. The rotary encoder and OLED now use the same wires and we plan on using them for the assistant box as well to avoid this issue in the future.

We also finished the startup code of our device which now includes choosing the language, deciding if the user wants braille or audio mode and entering the PIN. For the audio mode there is now spoken instructions for each function and speech recognition works with no internet connection but are working on adding that to the main script. Lastly for the rotary encoder we have edited the bounce time so there is less delay when scrolling through the letters now.

Week of 3/15/2021

This week, we worked on making a new design for the 3d printed box with a snap-able battery lid as well as making minor changes to it to fit the microphone and dc converter better as well as adding mouse ears at the corners to prevent 3d warping.

We also continued work on soldering wires to the uln chip that will drive the vibration motors on the glove. The necklace strap also shortens to chest area. It remodels with less electric tape but still no exposed wire.

We research wireless charging and found that inductor wireless charging might work for our product. Lastly, we completed the weather app where the user inputs a zip code and gets the weather back, found a new approach to face recognition, fixed problems with temperature sensor app and added a language selection option at beginning of main program.

Week of 3/15/2021

This week, we continued refining the pin entry for the assistant device. We also got speech recognition or speech-to-text working. And continued working on OLED display to be better in synch with rotary encoder. We have been working on getting the main script to work better and continuing work on pin the mechanism to make sure the springs don't hit each other when activated. We also completed the code to save pictures of faces under the given name for the face recognition library to look through.

We cover the wire from necklace with electric tape and secure it with copper wire. it will be hang over neck with vibration motor place on back of neck.

We done making cover for Glove and have doublecheck DC DC converter for 3V using multimeter. The value for amp and voltage are shown as following:

Lastly, we making demonstration of our Pin mechanism as shown in video:

Week of 3/8/2021

This week we got the OLED display to work in real-time when the rotary encoder so it's easier to see what you're typing out. The PIN code works better now and we now have it programmed where after 3 incorrect login attempts the device shuts off. Face recognition over the video stream from the necklace works now and we are working on having the TOF sensor work alongside it now. Lastly, we have been updating the CHP 3 section of the report and have generally been updating the wording/grammar in the other chapters and adding new similar types of products in order to better reflect what is the industry standard of our devices

We managed to complete circuit diagram of all component we use for glove and we will getting ready to wire and solder it next week.

We decided to start making our own conical spring at home using a bike or guitar wire. We placed a small amount of solder at the corner of the L-shaped pin to prevent the conical spring from sliding out and placed a PVC tube with a 0.25-inch slit on top of the electromagnet to prevent the pin from moving left or right. We then adjusted the spacing between the pins to 2.5mm and drilled the hole with a bench drilling machine so it met the requirements of the braille guidelines we are following.

Week of 3/1/2021

This week we have improved the communication between the devices where now the assistant device can ask a question and answer it with the glove to run certain apps. All wires are soldered to the board for the necklace and all the glove wires are soldered besides the vibration motors. Face recognition is close to getting done and the pin code entry script for the assistant device is completed. We have completed the preliminary testing of various scripts like the rotary encoder typing messages out which were sent to the assistant device, tested TOF, thermophile.

We have been working on getting the pin mechanism to work by using a spring to hold the pin inside the box when upside down where it sort of works with a custom spring, and have looked into alternatives in case it doesn't work like going to a machine shop/controlling servo motors. And have also looked into how we are going to keep the temperature inside the box below 70 degrees celsius as our 3d printed box will melt if we go above that.

Week of 2/22/2021

This week we worked on testing the three 18650 batteries in our battery holder with a multimeter to test the amount of output voltage we are getting to make sure it fits within the voltage range of the converter. After doing that we soldered the battery holder to the converter and hooked it up to the raspberry pi and ran it to see if there were any problems. We tested the UPS backup battery with Raspberry pi to see if it able to powered it when power adapter unplug.

We also have completed the model of the box we are going to print in 3d which comes with a battery holder and spots for the speaker and OLED.

We decided to try a rubber band for Pin mechanism due to its elastic properties to hold the pin inside the box when the electromagnet was off. We then build and tested it with 6 pins and found from the test that only 1 Pin managed to go up/down while the other 5 pins did not when the electromagnet turned on/off.

We also met again to connect the pin mechanism of the assistant device directly to the raspberry pi by GPIO pins and had to solder together 6 vcc wires into one wire and connect that to the raspberry. After doing that we ran various letters through our English to braille conversion code and found that it works but we are still having trouble getting the pins to work while being upside down. We also fixed problems with the mesh network where it now runs by adding static ip addresses to each device and have fixed the space, backspace, and send button on the glove. Lastly we added the heart rate and oximeter monitor to the necklace.

Week of 2/15/2021

This week we worked on researching a way to connect a battery backup with our assistant device as well as working on making our assistant box smaller and looking for the best batteries/devices to use in our device in order to take up as little space as possible.

We have also finished with the rotary encoder on the glove and it can type in english/spanish/french. We have also completed spacebar/backspace/send button on glove and it now works on OLED display.

We worked on researching how to connect grove magnets without grove pi to save space and give us access to necessary pin on the raspberry pi because we are running into coding issues in the grove pi with library functions not working on it. We have also worked on face recognition over video stream from necklace to assistant device and asking for user input from glove to control app features, and are continuing to work on heart rate sensor.

We also decided to change from a U-shaped copper pin to an L-shaped one which shortened the height of the pin mechanism to 1 inch which made it so it could fit with the other components inside the assistant box easier and likewise enabled us to shorten the assistant box to 2.5 inches tall so it could fit within backpacks or purses. We test it and it a little better but it still need to bring pin back inside. So, we add the compression spring from pen inside the tube but it not working at we expected unfortunately. We try to glue tube next to each other to make our dimension smaller but it very difficult to do so.

We try to make a block with 6 holes drilled into it instead of tubes to help the gluing process while also fixing the issue of the tubes being at uneven angles. We drilled 6 holes with an in-between spacing of 10mm on a plastic block using a bench drilling machine instead of a handheld drilling machine for a more precise vertical hole. After drilling each hole, we double-checked if the pins could go through the holes.

Week of 2/5/2021

This week we attach all components for the sensing glove and are in the process of testing it with the mesh network to see if it can activate the braille cell with the 6 vibration motors, along with the OLED and rotary encoder with our main program code. We also wearing it to see if it comfortable.

We have also found a way to fit all of the components of our main assistant box into a smaller pre made box without having to make an indented spot for the power bank.

We have also have a function for the TOF sensor, and thermopile that when running sends a signal to the necklace to collect data from the sensors and sends it to the main assistant device where it can be displayed.

We met up again to put all 6 of our magnets together in our assistant box to test all of them at the same time by using the measurements and spacing we confirmed last week. Unfortunately we confirmed that we are getting electromagnetic interference between our magnets which is attracting the other metal pins and stopping them from moving up and down through the pin hole. We have thus began testing to see if copper foil tape will help in repelling the magnetic field's interfering with each other and have also come up with a back up plan to use solenoids in case the magnets don't work.

During researching, we decided to use a copper pin because it is not magnetic, and we planned on attaching a magnetic steel object on the hook end of our U-shaped pin to minimize the amount of magnetic attraction between the electromagnets.

Week of 1/25/2021

This week we met up in the park to test 1st prototype of pin mechanism for the assistant box then found and measured changes needed to make it work.

We also used an electric meter to measure the current draw of all 6 of the electromagnets to help determine the power bank we need. The current and power of all 6 of the electromagnets are shown below:

Week of 1/18/2021

This week, we got the time of flight sensor to work and also got the peer to peer mesh connection between our 3 devices to work by enabling wifi direct without and can work without internet connection, this means we can now send text files, jpg images, and OpenCv between them and allows seamless communication between the 3 devices when sending braille messages etc. (major milestone completed). Also tested the pi sugar power source for the glove/raspberry pi zero to confirm it works. And soldered wires on glove, found that we need insulation for them and will finish soldering comp. on it Saturday.

Week of 1/11/2021

This week after getting back from Christmas break we worked on testing various components for our assistant device like the capacitive touch sensor and worked on troubleshooting the code for in order to get it to work. We also made progress on getting the electromagnet mechanism done by custom ordering a 3d made part for our magnets. And we met this week to solder our heart rate sensor for the necklace and are planning to test our power sources with an HDML cable next week. We've completed the multi language translation app as well as the weather app both through web apis, and begun writing the main program which will choose each app based on the users input. The speech recognition app is also complete and we are working on making it work as an input for the other currently working applications. We've begun testing the Time-of-Flight sensor for obstacle detection, however it's not giving proper outputs so we will rework the code, and possible test with a new Time-of-Flight device incase it is a problem with the sensor itself. We are also trying new methods of getting the Peer-to-peer mesh network to function offline for data transfer between the three devices.

Week of 11/1/2020 - 1/4/2021

This week we began looking for alternatives for creating a pin mechanism for our main assistant device and ended up deciding to 3d print the part we would need. In order to create the part needed, it was necessary to start learning a 3d designing program. We also worked on the weather app that will be included in the main software program of our code for the project that will tell the user the local weather based on zip code.

We also worked on getting the connectivity of our devices to work by researching how to connect them through peer to peer by creating an adhoc network on the main device which would assign ip addresses to whatever devices tried to connect to it, this however did not work.

We also worked on the code to convert text from an input file to braille display in the assistant device. Also the code to send text from converted image to text files from the necklace device to the assistant device is almost done, and we used LED's to use as an example of how we plan to turn the magnets on or off from the raspberry pi.

IGAD

⠊⠛⠁⠙

Progress Log