Posted on

Ziyu Li’s Status Report for April 6th, 2024

What did you personally accomplish this week on the project?

Two major things is done this week: 1. using hand to hold the motor to actuate the slider from last week, I identified that the best place to mount the motor is actually above the slider, not underneath like I expected. This is likely due to when motor is underneath, it also provides an upward force to the slider, causing it to jam with the pins more easily. Mounting the motor on the top can eliminate this non-horizontal force since the support for the slider is underneath. 2. Given this, I redesigned the whole actuator by allocating more space between the cover and the slider, I also designed and tested the motor mount, which at the moment does not work perfectly since their are some tolerance differences from the 3D printer, but I am confident that I can solve this tomorrow.

Is your progress on schedule or behind?

If compared to the original plan, which is to build 10x of these actuators and having with coordinate with each other, I am definitely behind. However, since the interim demo, we acknowledged that having one braille cell work perfectly is more important than building a ~70% working whole, and given the difficulties in this process is much bigger than we expected, we have scale down to that. And in this regard we are on schedule.

What deliverables do you hope to complete in the next week?

A single braille cell working, shifting to different braille patterns.

Test that I am planning to run

After I get the braille cell working next week, I’ll focus on three major tests to make sure our product is meeting the original design requirement:

  1. activation speed: the slider should be able to move from one end to the other  under 200ms.
  2. Jam frequency: the slider should not be jam in between at least 15 actuations (switching from one position to the next under the speed outlined in 1)
  3. Jam recovery: the actuator should automatically detect jam situation and recover from it using the back-and-forth motion in under 500ms.
Posted on

Yujun’s Status Report for April 6th, 2024

What did you personally accomplish this week on the project?

I researched on various web application making tools and decided to stick with flask for its usability and compatibility with python which is the language the translation algorithm is made on. So far, I’ve created a simple web page with a text box which can take inputs from the user and store the input data within the web page domain. Next week, I plan to create a fully function web page that is able to generate a prompt for the user to type in the input or upload a txt file, and is able to generate a braille encoding file from the input.

In particular, how will you analyze the anticipated measured results to verify your contribution to the project meets the engineering design requirements or the use case requirements?

The anticipated results for testing of the web application and braille translation algorithm is as following:

– Translation Reliability : close to 100%

– Data transfer reliability: 95%

Right now, I have already tested the translation algorithm on 5 iteration of testing, each on varying types (braille-contractable, non-braille-contractible, prefixed words, short words, long words.) For each type 20 random words were chosen from a word generator online. On every iteration of testing, the translation algorithm yielded a 100% accuracy which was verified through pre-existing verified braille translation websites. The comparison was possible through a temporary display function made for testing which translated the encoding to the computer terminal. Thus, it can be said that the accuracy of the translation met the translation reliability use case requirement, and we do not have to consider the ethical concerns for the project that pertains to translation inaccuracy and unreliabilty.

In the future, I will be testing the reliability of the data transfer between the web application and the braille pad. To be reliable we decided that if we input 100 words, at least 95 words should be able to be transferred successfully to the pad as based on a normal person’s learning curve, around 70 words and braille patterns can be fully memorized within a full school day at most. Thus, our 95% is sufficient for reliability. To test this we will iterate the transfer of 100 word inputs to the braille pad 10 times via wifi transfer to find the average. If the wifi transfer fails the test, then we will resort to serial bus transfer (higher accuracy) as a back up plan.