Possible FLR Hacking?

[Koko Ed]

Originally Posted by Chris is me

Second - why would they target you and not the teams in the run for winning the event?

We were seeded fourth so we were in the running for winning the event.
And apparently something fishy was going on. Our team leader said someone was in our pits and may have been messing with our bumpers and left a note on our robot . He contacted the volunteer co-ordinator to sort it out.

[Tuba4]

Don't discount the idea of someone remotely controlling a robot by iPhone /iPad. There are two free iOS apps called DSLite and IAMDriver which CAN drive a robot. It is not out of the question.

[pfreivald]

Hacking WPA2 seems tremedously unlikely.

Robots dying mid-match only to come back on after a while is likely because, unlike last year's radio, this year's bridge has a long reboot time for even momentary power losses.

Last year, a momentary power loss means a stuttering robot control. This year, it means a dead robot for most of a match.

[jtdowney]

Originally Posted by Tuba4

Don't discount the idea of someone remotely controlling a robot by iPhone /iPad. There are two free iOS apps called DSLite and IAMDriver which CAN drive a robot. It is not out of the question.

DSLite would still require the iPhone to have the wireless key and iAMdriver only works if you're using the iAMdriver control system instead of the cRIO.

Originally Posted by sparrowkc

WPA2 security, as used by FIRST, has no known weaknesses, meaning that there no methods better than brute force for obtaining a key. WEP has critical weaknesses which can be exploited by the software you mention, but FIRST doesn't use WEP, nor should anybody else.

Not true, within the last year a vulnerability was disconvered in the WPA2 spec known as Hole196. However it doesn't help in finding the key only helps you do nefarious things once you know it.

Also there is really no need to brute force the key, it would be far easier to just take the pieces of paper that have the wireless keys written on them. Many of the competitions I've been at teams will leave the paper they're given with the key just laying around their pit. In the frantic pace of the event no one would likely notice if it went missing.

All of that being said I have significant doubts that there was a malicious person at FLR remotely controlling robots. I've seen many robots do weird things over the years due to low batteries, bad wiring, faulty programming, etc. We build very complex machines and when they don't function as expected it can be frustrating. However there should be a lot of thought given before anyone makes accusations of cheating.

[BrendanB]

Originally Posted by Mark McLeod

Yea, 10 volts is a depleted battery, NOT a charged battery.

Don't EVER play with a 10v battery...

Agreed! At WPI this last year 1915 (1519's practice robot) lost a key match as I couldn't control the robot since the voltage was dropping down to 10 volts which caused erratic and uncontrolled behavior.

Also check your controls and code!

[PayneTrain]

While it's hard to not blame the FMS connection with the robot for being consistently pathetic as it has in years past (waiting for 3 hours in the stands to connect for a match), one unfortunate thing I have come to know in the last three years of FIRST is that not all of the thousands of teams go about competition the "right" way.

[mwtidd]

Also we found that before each match, the driver station computer should be restarted and booted up only in the driver account. this prevents any other programs from running. we has issues with the developer account which when started up didn't always shut down processes.

We saw very similar behavior to what was described here...


...maybe Anonymous is coming after FIRST...

[BrandonD-1528]

Originally Posted by Bongle

10 volts (which is really, really low btw) at idle could certainly drop to 3 volts when you start actually driving.

Here's an alternate theory:
-The 10V robot started the match, and as soon as the drivers engaged the power-hungry main motors, the voltage dropped so low the various voltage regulators couldn't keep the cRio and bridge functional
-The 2011 bridge takes about 90 seconds to reboot in our timing
-By the time the bridge rebooted and reconnected to FMS, the match was over.

I realize that 10 volts is low (someone on our team put the wrong tag on it), but it was a little unusual to see a robot lose connection right at the beginning and regain it immediately when the match ends.

To everyone else, I know the hacking theory is a bit out there, and it is possible FMS was being a total butthead of a system. On the first day of practice matches they almost made us switch to the 2009-2010 gaming adapters because the field failed to connect to the new ones for a good 8 hours.

Of course, it turned out that the wifi in the gym there was interfering with FMS.

And to anyone who was at FLR, I really hope none of you tried to connect to the "Free Publlic Wifi" adhoc network.

[Matt Krass]

Originally Posted by BrandonD-1528

I realize that 10 volts is low (someone on our team put the wrong tag on it), but it was a little unusual to see a robot lose connection right at the beginning and regain it immediately when the match ends.
........
Of course, it turned out that the wifi in the gym there was interfering with FMS.

Two points here, first, as mentioned earlier at 10V it would be very easy for the startup current of a robot coming to life to cause the radio to suddenly lose power, and the radios restart and reconnect time is roughly the span of a match. Its not unusual at all, it's somewhat expected. Plus a battery at 10V after robot start is probably very near dead and is giving you 'false hope' readings because its under very little load. I can understand why at first glance, this doesn't seem to be a problem, but I assure you it is a problem.

Second point, I've seen wireless networks with 20+ computers on them, actively operating in very close proximity for months at a time (ever see what happens on a college campus?) with no interference problems. The WiFi system employed by the field and robots should not succumb to the presence of one local ad-hoc network, it would never have been so successful in the household market.

Unfortunately I think at this point you guys have to suspect a problem on your end, I know it's not pleasant. I've been in the shoes of "it couldn't possibly be our fault, the field/wifi/arena/other team is messing with us" but I can assure you every single time the culprit has been less sinister. I urge you to take this experience and try to fine tune out any bugs in your system, and do whatever you can to make it more robust. This is the first contact for many of these systems and weird, new bugs are expected.

Matt

[XXShadowXX]

To everyone else, I know the hacking theory is a bit out there, and it is possible FMS was being a total butthead of a system. On the first day of practice matches they almost made us switch to the 2009-2010 gaming adapters because the field failed to connect to the new ones for a good 8 hours.

Of course, it turned out that the wifi in the gym there was interfering with FMS.

The volunteers at an event work with limited data. Calling them out for making you switch controllers is pointless. If you want to check every level of a problem do it yourself, if you want your robot working as quickly as possible cut some corners.

[Jonathan Norris]

Originally Posted by Mark McLeod

Yea, 10 volts is a depleted battery, NOT a charged battery.

Don't EVER play with a 10v battery...

This.

I am almost 100% sure this is the cause of your issues. Many of the components in the electrical system do funky things under low voltage, especially the jaguars and radios. Charge your batteries, you were playing on a dead battery which is never a good idea.

[Matt Krass]

Originally Posted by XXShadowXX

The volunteers at an event work with limited data. Calling them out for making you switch controllers is pointless. If you want to check every level of a problem do it yourself, if you want your robot working as quickly as possible cut some corners.

Uhhh... I'm not sure he was calling them out, I read it more as a "The field was acting up and they were trying some crazy things out of desperation". Also, cutting corners? I'd never advise a team to do that, it never actually saves any time.

Matt

[BrandonD-1528]

Originally Posted by Matt Krass

Uhhh... I'm not sure he was calling them out, I read it more as a "The field was acting up and they were trying some crazy things out of desperation". Also, cutting corners? I'd never advise a team to do that, it never actually saves any time.

Matt

I was indeed saying this. Their theory was that the 5v radios FIRST supplied were too underpowered to be reached by FMS, which I personally didn't believe. They ended up not switching to the old radios, because the problem was indeed fixed by RIT shutting down their wifi access points (there were like 20 of them in there). The volunteers did the best they could with what they had.

As for the battery thing, I'm aware that it was likely the cause of the issue. I just mentioned it because the robot didn't seem to power down, though I now realize the bridge probably died on its own.

And once again I mention that someone on our team mistakenly marked that battery as charged, and we didn't have time to check it before we queued.

[Matt Krass]

Originally Posted by BrandonD-1528

I was indeed saying this. Their theory was that the 5v radios FIRST supplied were too underpowered to be reached by FMS, which I personally didn't believe. They ended up not switching to the old radios, because the problem was indeed fixed by RIT shutting down their wifi access points (there were like 20 of them in there). The volunteers did the best they could with what they had.

As for the battery thing, I'm aware that it was likely the cause of the issue. I just mentioned it because the robot didn't seem to power down, though I now realize the bridge probably died on its own.

And once again I mention that someone on our team mistakenly marked that battery as charged, and we didn't have time to check it before we queued.

I didn't mean to sound harsh, sorry if I did, I just wanted to make it clear that as much as it might seem helpful to blame the evil forces of the field/etc, its only going to lead to more stress later. I hope you have a good season, and remember every time something breaks, it's a learning experience

[Bongle]

We were doing some testing with last year's robot today, and it's in pretty hurting shape. Even with a full battery, some very weird stuff can happen due to motors stalling. It is configured as a long-base 4WD tank drive with grippy wheels on front, slick wheels on back, and riding on carpet. Clearly, this configuration is really bad at turning.

Example of weirdness:
When we move the joystick full left or right, the robot sometimes shoots forward or backwards. We eventually figured that this was because one of the motors would trip its fuse while trying to turn, and the other one would then power the robot full-speed forward or backwards, depending on your direction of turn. So it ends looks like something un-commanded is happening, when in fact it's an interaction of the underpowered battery, fuses, and environment.