[RandomStyuff]

As the programming lead of an Israeli FRC team, and someone who was involved in fixing many of the communication problem (I stayed all night to help fix it, having found myself with my shoes not having left my feet for 40-ish hours) I feel I'm somewhat qualified to represent the other side of things in many of the subjects from the post above (-not all; things relating to the cost of things and the exact dimensions of the field are not what I worked on)

I will skip the first list of complaints as they talk about organisational complaints and I would not be a reliable source for facts on them.

Pit area

• Inspection process was held by inspectors who were not familiar with the Robot's rules or any other rules. This was one of the suggestions to improve the Israeli regional that I and other senior mentors suggested at the end of the 2009.
• As a result from this, many robots were illegal including robots which used motors not from the KOP.

My team did not use the practice field so I can't say anything about that. It seemed like it was enough carpet though, but I never really had time to go look. As for the inspectors, they seemed very professional and the inspector that handled my team seemed to know his FRCs and the rules, but I didn't have experience with all inspectors, nor illegal parts in an inspection, so it is possible.

Practice games

• Even in 2009 with less communication problem not all practice games were held in the first day, but at least every team was able to test the robot inside the official arena. In 2010 many teams were not allowed to test their robots on the official arena because of the communication problem. They simply did not manage to insert all teams by 10pm on the first day. Although they promised they will give practice time on the second day before the opening of the qualification matches many teams did not tried their robot inside the stadium.

The communications problems, which I will detail more about bellow, were the cause of this. There isn't much FIRST could do about it -teams barely got to play 3 games as it is!

• The management requested from six teams to leave their robot inside the stadium in order to examine the communication problem and fixed it. It seems a very good idea if all the teams were able later to test their robots and verify that they are working with the field system.

I was the one who went to request it from the teams. We used the robots to try the system under load and to simulate a game for testing the FMS (Field Management System). The fixes that we found were then implemented on ALL routers, one by one, and there was no advantage to having left your robot on the field. The only thing we did with the robots was test our fix! After that we went router by router and reconfigured them one-by-one!
I repeat: There was no inherent advantage to having left your robot on the field as part of the 6!

• I know that at least one team was able to test their robot from each of the six stations doing so for 2 minutes from each station.

While I don't know if this did happen (at 7 or so when teams came to get their robots from the field I went on my search for something with caffeine) it still doesn't give any teams and inherent advantage. Each station is pretty much the same when alone, and therefore they were just waisting their time!

The official arena

• Beneath the bumps there were two wooden surfaces that actually changed the structure of the bumps. This impacted severely the autonomous section of the game. These surfaces do not exist in the arena drawings.

While I am not an expert on the Breakaway Arena, I'm pretty sure that the wooden platforms near the field were perfectly legal and in fact part of the official spec. I recall an update having been made early on in the season which stressed the fact.
(On the rest I am not knowledgeable enough to answer)

Communication network

• FIRST Israel failed to use the (Access point) AP system that was sent by FIRST to Israel. According to the management it was not working at all and they knew of this problem three days before the practice day.

The AP supplied by FIRST this year were worse than those from the years before. We tried them during the night in many different combinations and settings. From what I understood, work on getting the FMS (Field Management System) to work started the moment it arrived. The APs supplied by FIRST this year were chosen because they were cheaper, not because they were better, at least according to our FTA from the USA. We saw no increase in performance using them.

• No site's survey was done prior to installing a wireless network inside the stadium. They could discover many wireless networks that cause communication interferes.
  

The Yad Elyuah stadium has free WiFi - That WiFi was disabled. The correct tools for wireless surveying are slightly grey-line on legality in Israel due to military purposes, yet we still checked for interference and found that while it did exist, it alone was not enough to cause the problems.
The stadium is not far away from the "Kiriah" - The top Israeli Military center- a Pentagon of sorts (and to all Israelis who are jumping up to say that it's not 100% accurate, it's close enough for the Americans to understand) which obviously also has sophisticated radio equipment. There is no real solution to this problem. You can't ask all of Tel-Aviv and the entire Israeli Military to stop working for three days or stop using their equipment just for a 'bunch of high schoolers with some robots'. Everything that could have been disabled, was disabled, and overall, interference was not the problem.

• When they discover the communication problems they did not invited ant experts to check it or suggest alternative solutions. Although they keep telling the teams that the best expert from Israel and online from the US working on it.
  

Now let me be a little bit more vocal on facts:
1) FIRST Israel invited a company with experts in the field of wireless networking to come help us. They came, and they tried fixing it. They didn't manage because the problem wasn't with the wireless!
2) At this stage, FIRST Israel invited an expert from the top Israeli Millitary's Intelligence group's (8200) Networking team. The guy is part of a team that has been credited as one of the best computer intelligence teams in the world, and is one of their elite networking experts. I think that would qualify as an expert from outside. The expert stayed with us all night long, and although he went to work or rest (not sure) for some period, he was back in the arena in time for the finals to make sure that everything was running properly.
I doubt he is reading this or will get this in any way, but he is:
It was truely a mindblowing experience working with you; I consider myself somewhat of an experienced networker and my understanding of networking nearly doubled over that one night. Thank you!
3) Throughout the whole time, we had experts from the USA, including the designer of the system himself with us on Skype. While they tried to be helpful, we always seemed to be a couple of steps ahead of them and they weren't much use in the end.

• Instead of a six AP (for each robot a cell with 300MB width) they used a single N AP meaning for all six robots the width was 150MB instead of 1.8GB.
• When eventually activate the system with one AP no security mechanism, were implemented. Any wireless phone or computer could access the network and caused the available width to be reduced.
• Ideas we offered to the management to use three regular AP and connect each one of the three to two access point bridge and creating three different cells with the assistance of security WEP (Wired Equivalent Privacy) & Pre-shared key mode (PSK, also known as Personal mode) were not implemented.

The problem was not with the WiFi; bandwidth wise, there was no problem at all! Follow me with the math:
Assuming absolutely no compression, and maximum resolution (640*480), a single frame from the camera is 900kb: The maximum FPS if I remember correctly is 15FPS. That means 13.5 MB/s MAXIMUM per robot on camera info. All other info is at worst half a MB per second (and I'm exaggerating here). That means 14MB/s/robot*6 robots (again, assuming all robots are using their camera at full resolution, with no compression and full FPS)=84MB/s - Well bellow the spec of 802.11n! This is also assuming all robots on the field have cameras at that full bandwidth, which was clearly not the problem at the Israeli Regional this year.
Now, even if we did decide that we wanted to go with a direct method- It is IMPOSSIBLE: The FMS is not build for that - it is built for a single router that communicates with the robots, and if anyone believes that they can write a new FMS during a regional, and test it to a level where teams won't try to crucify you like this mentor for example, then you are clearly either on something, or inexperienced in software engineering. Actually, you'd have to rewrite part of the FMS protocols for this to work which would mean changing the code for each driver-station and rewriting the dashboards.

WEP was not necessary after the steps we took which I will explain bellow- a laptop or cellphone COULD have connected, but because we disabled the DHCP server they would have to have defined a static IP which was in our subnet, something which I doubt anyone would do on purpose, and I doubt even more that it could happen by accident.

• Network experts that were presented as guests offered there assistance and were rejected.

We had an expert from the IDF (Israeli Defence Force), any more expertise and it would have just become disturbing and overlapping. It would have made us inefficiently redundant.
--
Now to explain some of the problems we fixed: (not all)
- The DHCP server was on at first, and the WiFi did start off as unprotected, although it was protected before the first robots tried entering the field. The problem this caused is that the DHCP tables filled up and the leases of IP addresses were far from expiring, which means that the router did not have the resources to manage the new IP addresses connecting to it. Disabling the DHCP server and cleaning the tables, moving everything to static IPs fixed part of the problem (this is the 1AM part)
- After that was fixed, we discovered that the FMS router and the robot APs were not on the same subnet class. The robots were correctly configured on subnet class A, but the FMS was somehow locked on class C. We tried to unlock the FMS's router to work on class A, but nothing seemed to work. At this point the expert from the IDF went home and returned with his router from home, and we used it for the rest of the testing and the competition. (This was approximately 4 AM, when my dad woke up having fallen asleep in the car, after I told him to wait for me because I'll "be there in a minute"...)
- Although I was so tired I was nearly out of it by then, I think that we also limited the bandwidth per connection to make sure that it wasn't that problem. No robots reached their maximum bandwidth.


Although this hasn't come up in this thread, but in others before it: some people claimed that bad programming was to blame. I disagree. I saw the code of many robots, my robot also can communication problems and I had 6 programmers go over the code line by line, of which one is a software engineer and one is a former software engineer which now teaches computer science. If something like bad code on one robot can crash the whole system, then something is inherently wrong with the FMS.

I will not respond to the whole MisCar disqualification story, I don't know what happened there with enough certainty, nor do I wish to get involved. As for the stadium, that is logistics and organisation, of which I had no involvement either.

I hope this post brings new information to light and explains part of what happened. I would actually take the opposite stance and praise FIRST Israel for doing everything within their powers to get things working, even when it cost them more than they could usually afford. (Thanks Alysha for offering me a hotel room, even though we never actually got to the going to sleep part of the night )