GompeiScout

GompeiScout, named after WPI's mascot Gompei the Goat, is a scouting software for FRC 190. Developed by Katy Stuparu, 2023.

What it does

GompeiScout 2023 allows scouters and drivers to easily enter match data.

• Autonomous
  • Number of cubes and cones scored low, mid, and high
  • Mobility?
  • Docked?
  • Engaged?
  • Comments
• Teleoperation
  • Number of cubes and cones scored low, mid, and high
  • Floor pickup?
  • Human player station pickup?
• Match/Robot Analysis
  • Short summary
  • Analysis
    • Describe the robot's collecting/scoring mechanism. How does it affect the team's ability to score points?
    • Describe the driver's ability. How does it affect the team's ability to score points?
    • What is slowing them down? (ex. collecting cubes from floor, lining up to score high, defended)
    • What is their strategy (offensive, defensive, nothing) and how effective is it in scoring points?
    • How well do they work with their alliance? Discuss endgame here.
• Driver Input
  • How successful was communication with this team?

GompeiScout 2023 also can display a detailed summary of these results for each team, including average values for all quantitative data and a list of all analysis collected.

Additionally, a summary of data for the teams playing in an upcoming match can be displayed. This includes average cones and cubes scored in teleop and auto as well as auto balancing data and short summaries.

The rankings for all teams based on quantitative data can be displayed. The quantitative data rankings are calculated based on how many points the team would have scored for their alliance on average during a match. See the FRC 2023 game manual for all point values. These rankings help scouters compare teams who perform similarly in terms of autos and scoring in teleop. This is not intended to be a pick list; the qualitative data gathered should be used in conjunction with this to make a pick list.

How it works

GompeiScout has two components: the web server/application and the database.

The web server/application uses Gunicorn and Flask. Gunicorn is a Python WSGI (Web Server Gateway Interface) server. Flask is a Python WSGI application.

The database is MongoDB, which stores data in the form of BSON (similar to JSON) documents. MongoDB can be run on a local machine to store data locally, or data can be stored in the cloud. Match data and driver feedback data are each stored in a collection. The advantage of using MongoDB for this project is to have a flexible data model that integrates easily with Python and Flask.

Both the web server and the database can be run in Docker containers. Docker is similar to a mini-virtual machine. It provides an environment for applications to run independently. This means that applications can be easily run the same way across various machines.

Scouters can access GompeiScout via a web browser with a wired or wireless connection to a machine where GompeiScout is running.

To run GompeiScout on your machine:

1. Install Docker Desktop. You will be able to manage or stop Docker containers with Docker Desktop.
2. Download the GompeiScout code, and cd into the GompeiScout directory.
3. Build the Docker containers.

   $ docker-compose up -d --build
   

4. Now, you will need to configure the database to work with GompeiScout.
   1. Start an interactive shell on the mongodb container. Then, log in as root (the password is in docker-compose.yml under MONGO_INITDB_ROOT_PASSWORD).

   $ docker exec -it mongodb bin/bash
   root@c84d9a66f7e3:/ mongo -u root -p
   

   2. Create a Mongo user that Flask can use.

   mongodb> db.createUser({user: 'flaskuser', pwd: 'fire23', roles: [{role: 'readWrite', db: 'flaskdb'}]})
   

   3. Create the matches and driver_feedback collections.

   mongodb> db.createCollection('matches')
   mongodb> db.createCollection('driver_feedback')
   

   4. Create search indices for the keyword team_number in the matches and driver_feedback collections. This allows for search queries by team number to be executed faster and more efficiently.

   mongodb> db.matches.createIndex({'team_number': -1})
   mongodb> db.driver_feedback.createIndex({'team_number': -1})
   

   5. Log out of root and try logging into flaskuser to make sure it is working.

   mongodb> exit
   root@c84d9a66f7e3:/ mongo -u flaskuser -p fire23 --authenticationDatabase flaskdb
   

5. You can access the website from localhost:5000.