FRC 250 - The Dynamos | 2024 DynoScout Build Thread |

Team 250 The Dynamos released our scouting code from last year, you can read more about the system as a whole here. During the build season, this thread will be updated about how our team approaches strategy.

From kickoff to Thursday our team was all together starting to break down the game and establish what kind of robot we wanted to make and on the 12th our software developers broke off and started the process to make our system.

The first thing we do is make what we call a data dictionary. This is a document where we think of all the things we want to scout. This is also a place where we give different variables a common name to avoid confusion as we code, the column name in the database, the format of those columns, and notes if needed. The stuff in red is from last year and was left as an example but will be removed as we fill out the rest in the coming days. This took up our Friday and part of Saturday.
2024_DataDictionary.xlsx (13.2 KB)

After this was made we thought about what we wanted to give the drive team as a report.


This is what we came up with, not a layout but just the information. The red dots are an average and the blue dots are a trend.

For the rest of Saturday and Sunday, we started to make the UI for the program. The students went through and made what they thought was a good design, presented, took another pass through, presented again, and then came to a consensus. The students decided to have three modes this season, Auto, Teleop, and Showtime (Usually endgame but this is not a defined thing this year so we are having some fun).

Auto has where the game piece was acquired from with an option to say that the robot just Drove into the staged pieces. Where the piece was delivered to with an option of Missed. Also where the origin of the shot was from the robot. We are also going to keep track of where the robot was set up, if they left the starting area, and if that team’s human player was in the amp.

Teleop has the same setup for acquired now without the flag option because we only care about that in auto. Same Delivered and Origin. We also will be trying to track when the amplification was started and when the first piece was delivered, along with tracking if co-op was pressed.

The Scouters enter showtime mode when the robot starts to climb or the match ends. We will have a way to track delivery from the chain, we imagine this will only be used for the trap. How long it took the robot to climb, the final status of the robot, what stage they were in, how many robots were on the chain when they went to climb, if they attempted to climb at all, if they were spot lit, how many shots to the microphone were made, a defense and avoidance rating, and finally an observation of what the robot strategy was during the match.

Match event is a list of things that can happen but not often enough to get more eyes off control (stuff like broke down, no show, passed center line in auto). Scouter is the name of the person scouting and Team No is the team number. This brings us to the end of Sunday and next up is contoler mappings.

4 Likes

With the holiday on Monday and snow on Tuesday we are working again on Wednesday and starting controller mappings. The students made a first pass and then presented to everyone what they came up with. After this, the students appreciated how hard these mappings can be due to the amount of data we want to collect. On that point, we may be trying to collect too much data and are toying with the idea of combining some different aspects. An example is combining a delivery from the subwoofer side and subwoofer front as just one location to the subwoofer. There were some other ideas but nothing is office yet and will update you when that kinds of choices are made.

3 Likes

We have come to an idea of what we want our controllers to do.

This is what we came up with and a formal document that goes into more detail will be posted later with an updated data dictionary.

After this was made we looked at our UI once again to make any changes that needed to be made. These changes would come about as we discussed the game more and changed our minds on certain previous choices. In Teleop we are keeping the flag for deliver, whether the alliance pressed the Coopertition or not we will be getting from The Blue Alliance. In Showtime We are swapping Defense and microphone labels as well as combining the stage and attempt and then moving strategy to the current attempt location. Next up we will be finishing the data dictionary update and then working on a presentation that will help teach the system to new people.

3 Likes

We started the day by calculating the max score as a fun activity and came up with 431 points. Next, we finished updating our data dictionary and made the formal controller mapping guide. Finally, we started making a presentation to help explain the controls to new people.

Group_Consensus_-2024-_Controller_Draft_1.pptx|attachment (561.9 KB)

The database tab shows an example of a few lines of what our database will look like. Note it is not entirely accurate but gives the right idea at this point.
2024_DataDictionary 2.xlsx (13.0 KB)

3 Likes

I forgot to talk about our pit scouting. This was a collection of all the different aspects we wanted to pit scout. The one thing we thought was missing was if the robot could go underneath the chain.

3 Likes

We have started coding within the last few days so here is our updated data dictionary and updated controller mappings.
2024_DataDictionary (3).xlsx (18.9 KB)

Group_Consensus_-2024-Controller_Draft_1(2).pptx|attachment (558.4 KB)

This is what our UI looks like when the code is running. We are currently working on the Auto code and will post a version once that is done.

image

image

image

2 Likes

After getting into controller mapping, we realized that fast switching on the DPad would sometimes result in the wrong input. So to fix this we swapped the Dpad and ABXY controls and switched the two joystick controls.

Group_Consensus_2024_Controller_Draft_4.pptx (588.8 KB)

2 Likes

We have removed the deliver origin variable because we can use logic and the current location variable to get the deliver origin. Also, we switch the Acquire and Flag controls because when one trigger is held down the other trigger will not work. I am guessing it is due to the library we are using but not 100% sure.

Here is an update on what our UI looks like, some changes have been made.

image

image

image

2 Likes

We have completed Auto mode and are now working on Teleop. Here is the link to our GitHub for this year: DynoScout24-Public

1 Like

Teleop and Showtime controls have been coded and are up on GitHub.

1 Like

A new version of the code is up on Git Hub. Now all the data is being sent to the database and the next steps are to reset all the variables, make the other 5 boxes, and then create the update database form.

This updated data dictionary removed some values we no longer use and added some new ones. The notes column shows what values should be recorded in the database. The database tab is a sample database so we can test out our match preview report. The Template and Lists Tabs are to make creating more database entries easier.
2024_DataDictionary.xlsx (30.6 KB)

1 Like

Some more database changes were added. We have a few more database things to add and some logic if a robot is a no show. Then we will create the update database form to help fix errors during use.

This is what our match preview report currently looks like. The tabs with TBA will pull data from The Blue Alliance. You will have to add your own connections for that to work.
2024_MPR-PUBLIC.xlsx (640.8 KB)

1 Like

Currently, all of the TBA connections are based off of 2023NYTR and the data manipulation within the queries would not be sensible or helpful to others. Once TBA establishes their match breakdown schema, the queries will be updated to a 2024 event. Once the MPR is finalized, the TBA connections and queries will be left in place with only the TBA API key removed so that anyone interested in using the MPR would just need to insert their key.

2 Likes

The next version is posted with some more database fixes. The update database form is next

1 Like

A new version is posted to the github. This version has some bug fixes and a start on the update database form. This weekend we were busy scouting week 0 to get everyone to learn the controls and find bugs.

2 Likes

Just an updated Data Dictionary.

2024_DataDictionary.xlsx (35.6 KB)

1 Like

We changed the controls to capture the center notes in auto to simplify things. Changes are up on Git Hub.
Group_Consensus_2024_Controller_Draft_7.pptx (543.0 KB)

1 Like

Some more bug fixes have been pushed

1 Like

Another set of bug fixes and improvements have been made. This time making the zone timers more accurate and some improvement in the center note logic.

1 Like

The newest version of the Match Preview Report is now available.

2024_MPR-PUBLIC.xlsx (1.1 MB)

1 Like