This week, Grape Expectations finalized our FDR submissions, including the video and poster.
We met with our coach to get feedback on both the video and poster, and incorporated the feedback into our current iterations of the submissions. We also met with our liaison, who approved our poster and video as well. We confirmed that our liaison would be able to attend our FDR presentation.
To finalize the poster, we made sure to include all of the content necessary to fully explain our solution. We also incorporated the feedback from our peers that commented on the poster on Teams. Mainly, we updated the colors for higher visibility, and altered the locations of some of the text boxes. For the video, we added subtitles, altered the noise levels, and added a few more clips to better explain the solution.
This week we continued our work to prepare for Final Design Review.
First, we filmed our video for FDR. Some of our team members starred in the video as a winegrower and a Terraview engineer, while others directed the shoot and wrote the script. One of our members then edited the video together with a recorded voiceover and additional representations of the Terraview UI and behind-the-scenes info about our project.
We also continued work on the FDR draft based on the outline we created last week. We started filling in information that was included in the outline. This draft will be continued next week and turned in for review before we turn in the final draft.
This week we began preparation for FDR, both in terms of the presentations and the report.
First, we started working on the draft of our FDR poster for the showcase portion of the FDR event. We started by prioritizing the most important images we wanted to share to represent our project. We decided that we needed images to represent the UI, the risk report, the data, and the ML models. These are the images we chose:
We also worked on the outline of our FDR report. We determined what we wanted to include in each section of the report, and are now prepared to write the draft of the FDR report in the next week.
This week was the Prototype Inspection Day, so we spent most of our time focusing on preparing our prototype and responding to feedback.
To prepare for Prototype Inspection Day, we wanted to present a client-focused presentation, since so many of our previous presentations have been entirely technical. We created a mock version of what the Terraview UI would look like with our new threats added based on the current Terraview UI. Using PowerPoint, we were able to simulate a user view of the Terraview system. Our PID went well, and we enjoyed speaking to all of the faculty.
After PID, we received the feedback from the faculty who visited our presentation. We were able to read and respond to the feedback in the latter half of the week. The main pieces of advice we received from the feedback were to explain our machine learning models more in depth. We were worried about our presentation being too technical, but it is important that users also understand what is going on behind-the-scenes in detail. We will be sure to incorporate more technical information in our FDR presentation.
As Grape Expectations returns from Spring Break, we continue to work on our prototype and finalize our plan in preparation for next week’s Prototype Inspection Day.
This week, we finalized all of the models that we are planning to show at PID. All of the code has been tested and packaged, and is ready to be integrated into Terraview’s systems. We also discussed the plan for what we would be showing at PID. Since many of our presentations have been very technically-focused, we wanted to do something more client-oriented and show a mock-up of the UI that the client would see when using Terraview’s system. Since we are not working on the UI in the capacity of this project, we photo-edited the base UI that can be found on Terraview’s website to give a mock-up.
In class, we were able to go over the technical needs of our project to prepare for next week.
This week the team worked to mostly finalize the prototype in preparation for next week’s break. In class, the team listened to a leadership presentation about leadership competencies.
For project progress this week, there were strides made on all fronts. First, the CS/BE team was able to get the API data scraper fully functional and Dockerized in order to be transitioned to Terraview’s team for integration. Additionally, the drought model was updated for improved accuracy in order to meet the TPM standards. All parts of the fire model have been fully Dockerized and are also ready for integration into Terraview’s microservice systems. All the relevant code has been pushed onto GitLab so that Terraview engineers can access it. Finally, rules for the pest model based on found databases and information were written in order to be used in a simpler model that Terraview already has. Overall, all parts of the project are on track and ready to be used by Terraview.
In class, we listened to a presentation from Dr. Denise R. Simmons about examining our own leadership competencies. We were able to identify which leadership competencies we were strongest in (and the ones we were weakest in), and how to use those to our advantage. We were also taught how to improve our weaker skills, as leadership is a set of definable, improvable skills that anyone can practice. This is important both in a classroom team setting, such as this one, but it will also be important as we continue into our engineering careers.
This week was entirely dedicated to our second Qualification Review Board presenation.
For this presentation, we had to demonstrate our testing plan and results, and our next steps based on these things. For the testing plan, we determined a specific testing plan for each section of the project: API, Fire Service, Drought Model, and Pest & Disease Model. We then presented our current project performance compared to our ideal Technical Performance Measures. So far, we are mostly on track to meet all the TPMs by the end of the semester, and we have made a lot of progress recently with the model accuracy score. Our next steps for the project are to further improve the accuracy and integrate more data sources, as well as to finish the API and continue work on the Pest/Disease model.
The judges provided positive feedback, and the main suggestion was to integrate client feedback and include ties to the end user to make it clearer what exactly our project will accomplish when it is fully integrated with Terraview’s current systems. We will definitely keep that advice in mind as we continue to work on the project. Now, to finalize prototype development in preparation for spring break and the Final Design Review.
This week the team made significant progress with our data scraper and model optimization, and learned about diversity, equity, and inclusion (DE&I) during the class presentation.
In terms of the project, the team reassessed the accuracy of our drought and fire models and worked to increase the accuracy to the level of the Technical Performance Measure that was set last semester. To increase accuracy, the CS team added new data sources, changed model parameters, and worked on developing new types of models to see which gave the best result when compared to test data. Additionally, the CS and BE team that is working on the API scraper continued development to find new fire data sources for the accuracy improvements.
During class, the team had the pleasure of listening to a DE&I presentation from Yvette Carter. During this presentation, the team was able to learn about our own implicit biases and how these can affect our performance and interactions on teams and in the workplace, as well as in daily life. Ms. Carter was able to teach us how to recognize and work to overcome these biases, a difficult undertaking that we will have to work for throughout our lives. It was extremely helpful to have these tools presented to us, and working towards diverse and inclusive workplaces is something we are all passionate about.
This week, Grape Expectations made more progress with all of the different facets of the project, and also learned about leadership in engineering.
In project progress, the CS team has made great leaps in the microservice work. They set up a weekly meeting with one of Terraview’s software engineers to work in the development environment and streamline the microservice integration. Additionally, the team has successfully Dockerized the code needed for the microservice, so it simply needs to be combined with Terraview’s systems in order to be complete. On the pest research side, significant strides have been made in the research that supports the continued development of a pest threat system in addition to the pre-existing wildfire, smoke, and drought threat systems. The main pests and diseases that will be studied are Pierce’s Disease (caused by Glassy-Winged Sharpshooters) and Powdery Mildew.
In class, the team listened to other teams’ “Design for X” presentations, and was able to gain a greater understanding of the Design for X system. Additionally, there was a guest speaker who taught about the importance of leadership in engineering.
This week, the team continued work on improving our prototype and also learned about the expectations for the Final Design Review.
In terms of work this week, the members of Grape Expectations have been divided into specialized roles based on what still needs to be done for the project. A CS and BE team are working on the development of an API scraper to find and integrate new data sources for the original fire model and the new smoke model. One of our other CS members is working on the integration of our current fire model into a microservice for Terraview. Another CS member is working on improvements to the drought model. Finally, the other BE member is researching and documenting pest and disease threat sources to evaluate whether a pest/disease model would be a feasible endeavor given the time left with the project. Everyone is full steam ahead and working towards overall improvement of the project.
In class, we learned about the concept of “Design for X,” a model used to determine what we will be designing for and why. This helps us keep in mind the end goal of our project while working on it, so that we can optimize the outcome. We have determined that our team specifically will be designing for testing and integration.