This week I completed my work on the Notecard. I connected the RPi and the Notecard together. I wrote some code which allowed me to upload a photo and metadata. This went pretty smoothly but I did have some issues with the upload issue speed initially. I worked with Kevin to debug any issues which were completely able to solve. Our uploads when timed were actually under 1 minute by a lot, even when underground. I then did some later work to deal with some Firebase issues where we were having duplicate uploads and our functions weren’t triggering. Next week, I want to actually connect the RPi to the car and make sure everything can work.
Dylan’s Status Report for 3/28/26
This week, I fixed the camera and inference pipeline to properly interface with the AI camera module, resolving integration issues that were blocking the end-to-end pipeline. After running initial integration tests on mock dash cam footage in Pittsburgh, I retrained the YOLOv12n model to improve detection accuracy based on observed results. I also assisted in integrating the upload pipeline on the Raspberry Pi 5, contributing to the completion of the full hardware-to-cloud data flow. I am currently on schedule, and my next steps are to integrate live GPS polling into the detection pipeline and conduct in-vehicle testing by powering the Raspberry Pi 5 from the car’s cigarette lighter socket
Team status report for 3/28/26
This week, the team made substantial progress toward a fully integrated end-to-end system. The most significant milestone achieved was completing the upload hardware-to-software-to-web-app pipeline: the dash cam now captures footage on the Raspberry Pi 5, transmits pothole detection metadata via the Blues Notecard over LTE, triggers a Cloud Function to receive and process the data, and displays the results live on the web app. This validates the core data flow of PotSpot across all three subsystems.
On the dash cam side, the team switched to the AI camera module as an interim solution, as the global shutter camera requires a lens to operate correctly. The software camera and inference pipeline were fixed, including a retrain of the YOLOv12n model for better detection accuracy, and the upload pipeline to the cloud backend was completed. Integration testing on real dash cam mock footage has also begun in preparation for the demo. On the web app, light and dark mode support was added, allowing users to select their display preference. The team is currently on schedule with the project timeline. Next steps include integrating live GPS polling into the detection pipeline and conducting in-vehicle testing by powering the Raspberry Pi 5 from the car’s cigarette lighter socket.
Kevin’s Status Report for 3/28/2026
This week, I worked on implementing/fixing the hardware to software upload pipeline. Additionally, I also deployed our web app on Firebase Hosting which can be accessed at this link here: https://potspot-487603.web.app/
Currently, the entire hardware to software to web app pipeline is working where we are capturing on the hardware and then sending the information via the Notecard by triggering a cloud function that would receive this information and then the web app would display this information. On the web app, I also added a light/dark mode where users can select their preferences. My tasks are currently on schedule with what we outlined and I am continuing testing on software with a main focus on hardware.
Dylan’s Status Report for 3/21/2026
This week, I focused on hardware bring-up for the camera module on the Raspberry Pi 5. This included verifying physical connections, testing the camera feed with basic capture scripts. I also began integrating the test footage into the existing recording thread of the multithreaded pipeline, ensuring frames are captured at a consistent rate suitable for real-time inference. Initial testing showed stable performance with no dropped frames under normal operating conditions. No major blockers were encountered, and next steps will shift toward integrating the GPS data stream into the pipeline and running preliminary tests from capture through pothole detection to Firebase upload.
Gio Status Report for 3/21/2026
This week I worked on the the Notecard. I was able to get everything registered on Notehub, so now I can sync from the Notecard to Notehub. Doing this I realized I had to create another Cloud Function that uploads on request which will trigger when I upload something to Notehub. I verified that the new Cloud Function runs which is good. Next week, I want to make sure that I can use the RPI and actually upload to Notehub. This would involve making sure the RPi5 can see the Notecard and then upload to it as well.
Team Status Report for 3/21/2026
As said in last week’s team status report, the most significant risk for our project right now is still the hardware. This risk is being mitigated as we are all working together to ensure the transition from software onto hardware is smooth. The contingency plan is to take more time on the hardware than the software right now. We are hoping to get the hardware components up and running so we can run preliminary testing. There hasn’t been any changes to the current design of the system. There also has been no changes to our schedule as well.
Kevin’s Status Report for 3/21/2026
This week, I updated bugs related to our frontend. I had a bug on the update implementation not updating on our Firebase Firestore data points. In order to fix this, I realized I had to create a new cloud function that runs when the “update” button gets pressed on the frontend. Additionally, I am still working on fixing a bug in our maps visualization implementation to accurately pin datapoints onto our map. I also did some more testing with different data points to see how our front end would react. This is on schedule with the timeline outlined in our design proposal. In the next week, I hope to start doing more testing and then finalizing the front end fully.
Dylan Status Report 3/14/2026
This week, I focused on finalizing and tinkering with the software subsystem architecture for the Raspberry Pi 5. This included refining the multithreaded pipeline structure, specifically the recording, inference, and upload threads, to ensure proper decoupling and performance on the edge device. I also reviewed the overall data flow from camera capture through pothole detection and local storage to the Firebase backend upload. No major blockers were encountered, and next steps will shift toward hardware bring-up of the camera module and GPS transponder..
Team Status Report 3/14/2026
The most significant risk for our project right now is the hardware I think. Since we have made a lot of progress on our software, so now transferring it to our hardware and making it work will be the biggest task. These risks are managed because we have thought a lot about how the different parts will work together. We also ordered different parts and have allocated enough time for new parts to come in the case where something doesn’t work. No changes were made yet and our schedule has not changed