There’s a scouting system called STAMP that was developed for this exact purpose in 2006; if you do a search, there are a fair number of threads about it.
Generally speaking, FIRST turns a blind eye to wireless LANs at competitions, seeing as they are pretty harmless (not to mention that the volunteer staff are way to busy to police that sort of thing). In the past, teams have run private wireless networks as well as networks open to other teams (STAMP, Sundial, etc.).
I know that STAMP was run at atleast 4 regionals as well as 2 of the fields at champs.
We will be starting to get a dev team together for STAMP v2 (2007) in about a week, so stay tooned if you are interested in joining! Hopefully now that it is a bit more established we can be even more sucessful then last year!
We’re hoping to do a fair bit with LAN and our scouting this year. Last year at Palmetto, we had a wireless LAN network up and a discussion forum on it, which we tried to get as many teams as we could on it. This year we hope to have a completely wireless scouting system, using Microsoft Access or possibly SQL or MySQL (apologies for the probable slaughtering of the spelling, I don’t use it, it’s just been suggested to me), where our scouts can go on their laptops (the school gives us wireless-enabled laptops) and fill out the form, and either live-update or periodical update to the main server (which someone, most likely me, will be watching). Then whoever is watching the server will compile the scouting report for the next match and send it down to the drive team. How publicly available we make our scouting system is yet to be seen, but we should have it ready in time for the NJ regional.
I have never seen rules regarding wirless LANs before.
We used one back in 2005 at the Arizona Regional; a scouting crew in the stands used a program to enter information, then submitted that information to a main file we had on a laptop in our pits. It also allowed us to use instant messaging between the stands and pit as a form of communication, instead of having to run a person up there everytime we needed someone or something.
We allowed other teams to view our data, but you needed the program we wrote to view it properly, so I don’t know how many actually used it (we gave a few copies of the program out).
I would love to use this again; paper scouting is less prone to loss, but is (in my opinion) a lot more time consuming to extract information out of.
We were thinking along the lines of:
-Setting up a wireless LAN
-Giving out a password to each team
-Once the team has filled out all the information about themselves, they are allowed to see information on everyone else
-Teams are allowed to edit their own information, along with a ‘second’ team page for everyone else
Of course, this would be like a wiki, where its all based in the team’s truthfulness, and judging by the amount of GP, i doubt anyone would lie about their robot (besides, if we get enough using it, and they lie, that can only hurt them, right?)
We had that problem at Palmetto. Most every team had a laptop, but they only had one (for programming), or the laptops they had didn’t have wireless access. Maybe, if you wanted to, you give out wireless cards under the assumption that because of GP, the teams will give them back at the end of the competition.
And having it advertised, either by flyers or by loudspeaker announcements would get the word out very well, I would think.
Building on this, you might want to try to contact all the teams before the competition to inform them you’ll be offering this scouting data; just ask your Regional Planning Committee for a contact list.
(i was going to start a new thread for this, but its all related to this scouting project)
Has anyone found anything on ranking algorithms… something like college football uses, without the 1/4 subjective part?
I would like to make some sort of recursive program that ranks teams not only on their record, but also factors in their alliance, the alliance they played against, and the score of the matches. This might take several months to perfect, and starting from scratch isn’t exactly my idea of fun. (on a side note, the algorithm doesn’t have to be quick, it won’t run anything more than 90-100 teams at a time, and if its programmed in C/C++, it’ll be fast enough)
You are describing exactly what STAMP (Statistical Team Analysis of Match Performance). The site’s domain is down at the moment (stampscouting.org) but will be up in 24hrs. We use a number of ranking algorithms based on just about every element of the match.
STAMP allows you to create your own ranking system. Pretty much you “weight” each catagory at a certain value so you can create your own summary of how you think they should be organized. I will try to dig up a screenshot of the v1.5 ranking page so you can get a better idea.
Not sure what you mean by sudo code for it. Feel free to take a look at the ranking page below. That is in the code for the v1.5 release.
When we first started creating STAMP these questions were discussed. The conclusion we came to is that this user should be left up to the user, not the developers. That being said, I can give you some more information that might better the answer.
What should be apparent: someone has to enter this data. The best way to do it is with a team of 6 scouters, one following each robot. What people DO with this data is up to them. Some teams may choose to make it public, others may not.
As for working with other teams, feel free! STAMP is set up with user accounts just for this purpose. Talk with other teams that will be attending your regional and create a joint scout team! We know that 6 people is a pretty hefty amount to spare from one’s team, so get together with 1 or 2 other teams and your down to 3 or 2. Just as an example, 492 worked with 93 at nationals on Newton to gather data.
Pretty much it all comes down to the user’s choice. We are not going to say that you MUST share. Sharing is great, it is certainly GP and if that is what you are shooting for then go right ahead. We see the STAMP dev team as just that, a dev team. It is up to you to choose how you use it.
In that vein what regional(s) is/are your team attending? If you are headed over to the PACNW in Portland we could certainly work something out. When STAMP v2 starts going into development I expect people to post threads in the scouting section about hooking up, so if you are not coming to PACNW and are interested in working with another team then feel free to be the first.