Social recovery: Research and education
Project brought forward by partner organization: British Columbia Institute of Technology (BCIT), Remotely Piloted Aircraft Systems team members Dr. Eric Saczuck and Sean Galaway
Eric and Sean have utilized Drones and Remotely Piloted Aircraft Systems to capture aerial images of various locations. They have kindly shared a subset of the original dataset with us for the DataJam and does not contain sensitive information. However, images may not be used for a purpose other than building the database. Images are not to be studied nor distributed outside of the Vancouver Datajam. If you are interested in the drone image data set or want to learn more about their project, please get in contact with their team at BCIT's RPAS hub website.
The goal of this project is to develop a data ETL pipeline using Python on AWS to organize images taken by a drone. Participants will receive a set of images, and tag information, and the goal is to develop a database that allows others to query images by date, tag and location. The team will focus on building an MVP (Minimum Viable Product - a product with just enough features to be usable) in the hackathon.
Using cloud is a nice way to ensure that the MVP built during the hackathon can be developed further as required and allows easy scalability (store TBs worth of images) that is important to the data owners.
This project would be done using AWS cloud services. Python will be used for scripting purposes as required. The data will be stored using S3 and the metadata will be stored using DynamoDB. S3 is an industry standard way of storing data on AWS, and DynamoDB is a NoSQL database on AWS. Boto3 is the Python library that we will use to interact with the AWS services using Python.
This project provided participants an opportunity to develop skills in the following areas:
- Backend development
- Database and system design
- Data engineering and ETL pipeline development
- AWS services usage (S3, DynamoDB)
Team lead: Madana Krishnan V K
Mentors: Luciano Andrade, Dr. Todd C. Brett
Team Members: Xinyue Chen, Annika Wevers, Enwongo Ekanem, Ruiz Rivera, Arisa Yoshikawa, Niranjan Ramesh
Main page: https://vancouverdatajam.ca/
| Date | Action item |
|---|---|
| Sep 13 - 17 | Participants are let in Discord, teams are formed |
| Sep 18 | Workshop day! Keynote: Making AI responsible with May Masoud |
| Sep 19 | Project statements are released |
| Sep 19-24 | Teams may work asynchronously (limited help desk support) |
| Sep 25 | Keynote talks, help desk support provided during the day, project submission deadline, career panel. See speakers |
| Time | Action item |
|---|---|
| 8:00 - 8:10 | Land acknowledgement, opening remarks |
| 8:10 - 8:40 | Keynote: Role of Statistics in Data Science: Applications in Biomedical Sciences with Prof. Jemila Hamid |
| 8:40 - 9:10 | Keynote: How to use the tools of data science to benefit Indigenous peoples and organizations with Hannes Edinger |
| 9:10 - 9:30 | Keynote Q&A |
| 9:30 | Help desk opens up, teams work on their project |
| 9:30 - 10:00 | Teams brainstorm tasks for the day |
| 12:30 - 13:00 | Team check in: share exploratory analysis results |
| 15:30 - 16:00 | Team check in: teams discuss presentation format and preliminary results |
| 16:00 - 16:45 | Teams prepare their 5-10 minute presentation, teams ensure all code is documented and stored in GitHub |
| 17:00 | Project video submission deadline |
| 17:30 - 18:30 | Project videos released on YouTube. Vote for your favourite team demo! |
| 18:30 - 20:00 | Career panel |
| 20:00 - 20:30 | People's Choice Award presented. Closing remarks |