Weekly Minutes with Professor Notash
Week 26: Tuesday, November 8th, 2022 (5:00PM - 6:30PM)
The team showed Professor Notash that the popups for the "No Mask Detected" and "Temperature Too High" were successful with video demonstrations. The team also proved that the calibration of the Pimoroni camera was successful at 2.5 feet; this was shown with multiple tests performed by different subjects.
​
Professor Notash asked the team to record more videos that accurately depict what is going on; the team will need to record the auto homing again showing one of us in the field of view of the camera; the team also needs to record the popup windows again with one of us in the field of view of the camera to prove that the modified code works in "real time".
​
The team was also advised to complete the report by Thanksgiving Day, and to provide a rough draft on Friday, November 18th.
Week 25: Tuesday, November 1st, 2022 (5:00PM - 6:45PM)
Professor Notash gave the team advice on what to include in the report; It's important for the team to specify that the camera tracking does not utilize Logitech's software. The servos and the mask detection camera work together using TensorFlow and OpenCV.
​
Professor Notash also wanted to the team to add on to the outside Raspberry Pi GUI; when a person has a temperature that is too high, a popup must appear to inform them why the door is not unlocking.
​
Due to the team experiencing large margin of errors for the thermal detection camera compared to a hand-held thermometer, the professor asked them to do more test comparisons with the Pimoroni camera directly on the forehead; this may make the comparison more accurate because the hand-held thermometer was used directly on the forehead. Professor Notash also suggested placing a hot towel on their foreheads to simulate a fever. If the team finds that the distance is the reason for the large margin of error, they would have to change the design so that it detects the temperature on the back of a person's hand rather than the temperature of their forehead.
Week 24: Tuesday, October 25th, 2022
Professor Notash and the team did not meet this week. This allowed the team to allocate this time to focus solely on the project and ensuring the entire system works together.
Week 23: Tuesday, October 18th, 2022 (5:00PM - 6:00PM)
Professor Notash gave the team some suggestions on what to include in the report; It is important to include a disclaimer that individuals have different temperatures and everyone is affected differently by environmental factors. Children, for example, naturally have higher body temperatures.
​
The team clarified to Professor Notash that the system would be implemented for smaller facilities and that the door leads straight inside the building. This would not be used as a second door inside a building.
​
The team was asked to create another recording showing the magnetic lock LED when it is locked and when it is unlocked. The audio for the outside Raspberry Pi GUI should also be modified to inform people to wear a mask at all times when inside the building.
​
The team informed the professor that the pan and tilt system only follows one person at a time. The current problem with the two cameras is that they are unable to detect a person when it is backlit. The team has to test the system again with various lighting conditions and include backdrop to mitigate the issue of non-detection when the cameras are in direct sunlight.
Week 22: Tuesday, October 11th, 2022 (5:00PM - 6:00PM)
Professor Notash asked the team to change the disclaimer at the end of the COVID-19 questionnaire to avoid appearing rude. He also asked the team to redo the voice recording to prove that the voice communication actually works in real time. This would require Jon and Charlie to both be in the same video or record separate videos and edit it as a picture in picture video.
​
Professor Notash informed the team that he has been in contact with the panelists and was concerned about the team not matching the schedule on the timeline. He asked the team to push for 90% completion by Tuesday 10/25 with the entire system working together. After 100% completion, the team should focus solely on the PowerPoint Presentation, Website and Report. The presentation will be after the Thanksgiving break.
​
Professor Notash approved of the idea of reducing the pan and tilt system from two to one. This means mounting both the thermal camera and mask detection camera on the same servos. To move forward with this idea, the team has to change the 3D rendering of the system.
​
From here on, the meetings have been changed to Tuesdays at 5PM.
Week 21: Tuesday, October 4th, 2022 (5:00PM - 6:00PM)
The team presented what would have been the previous week's presentation plus additional updates they were able to add after the hurricane. Professor Notash suggested adding a question to the list of Covid Questions that asks the patient to agree to the terms of keeping their mask on indoors. Professor Notash also suggested creating a record of the questions answered, though this may be time consuming to implement in terms of synchronizing the questions for each guest seen on the camera.
​
Professor Notash also suggested contacting Professor Jerry Reed to ask for advice on the thermal camera as well as the speech recognition for the voice detection portion of the Covid Questions.
​
The team was informed that they would be given an extra 10 days for the project due to the delay caused by Hurricane Ian.
Week 21: Saturday, October 8th, 2022 (10:00AM - 11:00AM)
The team had a Zoom call with Professor Jerry Reed who gave them advice on the current state of the project. He informed the team that the Pimoroni MLX90640 had too low of a resolution to be able to utilize with Tensorflow. He suggested that including a backdrop such as a large Styrofoam board would mitigate the ambient heat from Florida's hot weather. Otherwise, it would be best to tell the patient to move closer to the thermal camera for a more accurate detection.
​
Professor Reed also informed the team about a Linux Speech recognition software that could possibly be utilized. He agreed that using the mask detection camera to move both both servos to point at the correct location. He also told them that he would look into the Vidstream package for voice communication.
Week 20: Tuesday, September 27th, 2022 (4:00PM - 5:00PM)
Cancelled Due to Hurricane Ian
Week 19: Monday, September 19th, 2022 (5:00PM - 5:45PM)
The team presented each of their work from the past week. Charlie showed how he programmed an "Unlock Door" button that would be for the GUI of the indoor Raspberry Pi. Because Charlie didn't have the relay or magnetic lock with him, he connected LED lights to the Raspberry Pi to show that the concept worked the same way. Jon showed that he was able to program the two pan and tilt servos. Though they did not rotate a full 180 degrees in the video, all that is needed for the project is 90 degrees. Professor Notash's feedback was to measure the weight of each camera to see if the servos would be able to support the weight of each camera. Charlie's job was to send these measurements to Jon and for Jon to test the weights on the pan and tilt servos.
Week 18: Tuesday, September 13th, 2022 (4:00PM - 5:30PM)
Professor Notash informed the team to test the thermal camera from further away (3ft) and determine if other objects/people will interfere with the results. The focus should be on the forehead; see if the person needs to go close to the camera to avoid ambient temperatures. Also see if TensorFlow could be utilized/trained to look for foreheads. He encouraged the team to test different mask colors to see if the mask detection camera is affected. Another test for the mask detection is in a "real world" environment where people are moving in the background. A screen recording of this would be beneficial.
​
The team needs to adjust the Power Budget on the Senior Design Project Tab and possible include the camera's fps to see if current/voltage is increased. The team also needs to add to the learned lessons, add code to the "Code" tab and start adding presentations again. The publications tab could be used to add the tutorial pages.
Week 17: Tuesday, September 6th, 2022 (4:00PM - 6:00PM)
Professor Notash discussed the type of sources that the team used and informed the used of IEEE Ethical standards. The team of course did follow this and will be including the references for the source code that was used for the mask detection software. The team asked if the secondary screen (inside the building) could be a regular monitor, mouse and keyboard set up rather than the 3.5" touch screen. This would allow for easier control over GUI and make it easier to read. The TFT 3.5" display also needed a stylus, so this would reduce the risk of losing/needing to use a stylus.
​
The team discussed their plans for the rest of the week prior to the next meeting. Jon will be continuing to test the magnet and connect a relay to the Raspberry Pi. He will troubleshoot the GPIO for the relay because it is currently only outputting 3.3V. The power supply needs 5V in order to send a signal to open and close the door. Charlie will start working on the thermal detection system once he gets the thermal camera from Jon. In the case that Charlie and Jon do not meet this week, Charlie will instead work on the VOIP/Intercom system. An existing application such as WhatsApp could be investigated and used as it is a secure application and the facility would have a secure firewall.
Week 16: Monday, August 29th, 2022 (5:00 PM - 6:00PM)
Professor Notash provided some inputs that the team needed to address. This included adding to the 'Lessons Learned' tab and moving this over to a section called 'References and Resources', which would also include 'References' and 'Other Publications'. He also said to add a section in the Appendix for the coding using for TensorFlow as well as OpenCV. He informed the team that the system criteria should reflect the system working together, NOT each of the individual components working individually. For the structure currently built, he informed the team that it was okay to leave the structure in the garage, despite the humid temperatures that may be expanding the wood; so long as the door does not become too stiff for opening or closing. He advised the team to use a 120VAC extension cable to power the system. Professor Notash also gave Charlie a warning to add his work to the Progress Log to avoid losing points.
Week 15: Monday, August 22nd, 2022 (5:00 PM - 6:00PM)
Official Kick Off Meeting for Senior Design (First Day of Fall Classes)
​
Professor Notash covered pertinent topics. We covered in detail the Design Implementation Rubric and how the grading scale worked. We had asked questions related to how our project physically looked so far and going over this rubric helped rectify some issues we had. We also covered the Design Project Report. Professor Notash made sure that we understood that some of the chapters were to be changed around and the importance of the new Chapter 3. We had also set some goals for us. For example, we need to test the magnet assembly prior to installing it. This is key so that we do not destroy the structure and the door by drilling into it multiple times. We also want to check the view angle of the 7inch screen at static level. This is key so that if looking at it from a low or high height it can still be seen with clarity. Once we start to edit and make changes to the senior proposal paper, we need to send it to the English department so that they may look over our report as we make the changes.
​
​
Week 14: Monday, August 8th, 2022 (4:00 PM - 4:45 PM) &
Tuesday, August 9th, 2022 (12:00 PM- 1:55 PM)
Introduction to Senior Design Project
​
Professor Notash gave the group a head start on what to expect for the Senior Design Project. He went over the entire Syllabus with the group to show the grading criteria as well as the target due dates. Professor Notash also showed the group the expectations and rubric for the project report as well as the implementation of the project; both of which are worth 40% each of the total grade.