Thank You IFI

[Natchez]

IFI, the Robonauts would like to thank you for your dedication and relentless field support at past FRC events ... WE MISS YOU! For over a decade, we took it for granted that someone would be on the field, wearing out the knees on a pair of jeans, making sure ALL of the robots were ready to go when we heard 3-2-1-GO!

Thank you from the bottom of our IFI controllers,
The Robonauts

[Tom Line]

Definitely thanks to IFI. It was remarkable how bullet-proof most of their systems were.

I'm also going to say, thank you National Instruments! The new controllers have capabilities the old ones never dreamed of, with many basic features built in that have allowed less experienced teams to address the challenge this year rather than learning just how to make the robot move.

In addition, the wireless uploads and data feedback that we get to the dashboards are incredible, and have helped immensely with debugging.

We all know that the last time a new controls system was introduced there were a lot of field problems - we know NI will get this system sorted out and running the same way IFI did.

I look forward to the day when the NI system is optimized enough to allow video streaming through the web camera (without rebundling, etc).

Perhaps this post should be a universal sponsor thank you - to past, present, and future sponsors that let us do this thing that we love.

[R.C.]

I wanted to start this post also but never got around to. Thank you IFI for all the support and thank you NI. Great stuff and thanks for investing in FIRST Robotics.

[lynca]

Originally Posted by Tom Line

Definitely thanks to IFI. It was remarkable how bullet-proof most of their systems were.

I think there is a big difference when one company is responsible for all communication systems involved in the competition.

I feel like we have transitioned from an all-in-one system (similar to Mac.) to a multiple vendor system (PC) where communications and components are all split into different vendors.

I don't really like those P.C. vs. Mac commercials but this one had some great robot action so here it is...
http://www.youtube.com/watch?v=uLbJ8YPHwXM

[Akash Rastogi]

Also thank you IFI for those oh so reliable victors! Not to mention VEX.

MORT <3 VEX

[Greg McKaskle]

In DC, it became apparent that there was some confusion as to where the different elements of the control system come from. I won't try to identify every sponsor/supplier, but I would like to point out that the HW elements provided by or designed by NI is the cRIO and the modules. That is possible in part due to the gracious parts donations of many NI suppliers. Some of them are listed in this article http://outsideinnovation.blogs.com/p...al-instru.html.

On the SW side, NI provided LabVIEW. WindRiver provided the VxWorks tools and IDE. The WPI SW libraries were developed by a team of people from NI, BAE, and WPI. Documentation was provided by this group and FIRST.

Support is being provided by a huge group of people from many companies and individuals. This can cause confusion as everyone is simply trying to diagnose issues and get to the bottom of things. NI is proud to join with other companies and supply what we feel are world class products to be used in the FIRST competition, but we do not deserve credit for running the events or for elements provided by other suppliers.

Good luck, and hope to see you in Atlanta.
Greg McKaskle

[Qbranch]

NI: I know you are pouring a lot of money into FIRST and I thank you kindly for it, but, well, I guess this is sort of a reminiscent post, and without this paragraph would probably sound unappreciative. This is certainly not the case.

IFI:

The three years I was in FIRST as a high school student were the final three years for the IFI robot controller ('06-'08). I had a FANTASTIC time learning about and writing code for the controller; the use of a Microchip (thanks microchip for all your money, effort, time as well) embedded processor added a great learning experience to robot construction for programmers. I know that the skills learned doing that close-to-the-machine coding necessary with the IFI RC proved useful for many who have since graduated from their respective teams. That is, at least it has been very useful for me. I'm still using Microchip processors for projects here at the U of I, at the moment to build an integrated control/monitor/datalogging system for a pulling tractor. Not sure I'd be able to program something like that up without all the experience I gained from working with the IFI RC. Also, in this age of multi-gigahertz processors, a little reality check about how much you can do with 40 MHz is never a bad thing.

Victors I believe will live on for many years to come. Perhaps not in FIRST, but certainly in other robotics and process control applications. For the price they are (I won't hesitate to say) the best general purpose motor drives one can buy.

Thank you, IFI.

-q

[Kyle Love]

Thank you IFI! I miss seeing the IFI guys at the competition that I knew by name or by face. IFI's devotion to FIRST got it to the high caliber that it is today. Some say that the control system was underpowered, but if it were not for them, we would have still be using an array of electronics and not be on the same playing field. IFI truly is a class act company, with great employees and GREAT customer service. I will go on supporting VEX and using word of mouth to add to your guys' amazing "street rep". Thanks again IFI! You will be/are missed.

[Natchez]

IFI you made it look so easy ... thank you again for the early years! We have now completed the 3rd year of chronic robot bypassing. For those not familiar with the history, here is a self-reflection CD post from March 8th, 2009.

Originally posted by Natchez

I was "drug" (okay, not drug ... a gentle pull of the shirt) off the field late Friday afternoon at DC while protesting the starting of a match with two robots that had the dreaded "NO COMM" blinking light.

To fill in the story for everyone, there were several times on Friday, easily more than 10, that Paul started matches with a robot not having COMM. It was a little heart wrenching because every time they started a match with a blinking light, we knew in the stands that the robot had no chance to come to life but the drivers kept hoping and you could see their hearts drop as the match went on. More than once, I heard the field staff blame it on the teams saying, "We can't fix their problems for them." When I did my stay-on-the-field-until-everyone-had-COMM protest, two robots (our robot had COMM), one on each alliance, did not have COMM; they had announced the teams and were just waiting on me to leave the field to start the match. Magically, without any of the team members touching their robot, COMM for both of the robots were established (i.e. it was NOT the teams problem ... it WAS a field problem) and we played the match with 6 live robots. Trust me, if there would have not been this problem all day long, I would have never been smart enough to know what was going on in our match.

Onto what to do - The Lone Star committee has discussed this issue and our MC will not leave the field until all 6 lights are dark or until a somewhat exhaustive effort has been made to correct the NO COMM problem ... just like the old days. We see it as completely wrong to pay $6,000 then get one or two of seven or eight matches taken away because of field problems; that's roughly $800 a match. Not only taken away from the NO COMM team but also from their alliance partners; potentially squelching a chance at a spot in Atlanta (okay, that's a little dramatic but it can have serious consequences). I know, it's not about the matches, it's about coopetition, BUT if we asked teams to solve the same problem and don't bother showing up to an event, I'd guess we'd have about 10 teams 'coopetiting'.

Another thing that we've discussed doing is to have the pits stay open exactly the amount of time that the field runs over on Thursday & Friday. The main reason is that many of our teams only attend one regional and many of our veterans are spending a significant amount of time helping other teams. If we do run over, this will give teams the expected amount of time to help others AND the allotted amount of time to get robots inspected and repaired. Although we think this is a good idea, we must coordinate with the facility and FIRST to make it happen. We're not too worried about volunteers because we would only require the core volunteers to stay longer ... and we're definitely up for the duty if it is good for the teams.

After three years, we should be discussing the few robot bypasses and communication losses per year, not the hundreds per year we are now discussing,
Lucien

[Tom Line]

Frankly, I'm shocked to hear it's this bad. I've been to many district competitions in Michigan, and I think I can count on one hand the number of times a match was started with a robot not communicating.

In fact, the field FTA at West Michigan announced flat out that they would NOT be starting a match with disconnected robots. He was true to his word, and they still managed to finish qualifications early.

I wonder if we can ask FIRST to keep statistics so we can see what type of problem we're talking about.

In many cases, for instance the Beach Bots, the problem was not a field problem. It was a hardware problem that was nearly impossible to diagnose. It's pretty hard to distinguish between the two.

[lynca]

Originally Posted by Natchez

IFI you made it look so easy ... thank you again for the early years! We have now completed the 3rd year of chronic robot bypassing.



After three years, we should be discussing the few robot bypasses and communication losses per year, not the hundreds per year we are now discussing,
Lucien

I have also been sad to hear the many troubled stories this weekend of robots losing communication to the field. FIRST as a community needs to look carefully at how to simplify the control system.

FIRST had a request for proposal process for speed controllers and the control system in 2008, I wonder if it is possible to restart that proposal process to allow the community a chance to build a better control system.

I look forward to the day when robot communication failures will be a rare occurrence at a regional.

[Jon Stratis]

Up in Duluth two weeks ago comm issues were pretty rare. There were a couple of teams that had recurring issues (and we worked with them to try to resolve those issues, and in some cases succeeded), but there were many more teams that went out and ran every match without a single problem.

There are really two issues with having a control system: capability and support.

First, capability:

The old IFI systems were great, but they lacked in capability compared to what we have today. Back in Rack 'N Roll we tried to get vision tracking to work, and had a horrible time with the camera that was available then. With the new system, vision tracking is actually pretty simple... but it brings with it problems some teams encounter, like having too high of a framerate and bogging down the CPU.

Back then, you had to use Spikes to control all of your 12V solenoids. Today, you have the additional option of using 24V solenoids, AND don't need to control them with spikes.

Back then, power management was interesting. You had multiple distribution boards for high and low current applications, you had DIN rails feeding those, there was no visual feedback when something was wrong, and there were no regulated outputs. Today, the power distribution board has all of that and makes everything dealing with power much simpler.

Back then, you only had Victors. I know, a lot of people still prefer the Victors, and that's fine. But the Jaguars bring a whole host of increased capability, and I know our team has found ways to use all of that capability over the years.

Back then, you could only use serial connectors for your driver station unless you bought a special converter to allow you to use USB.

Back then, you had channel overlap problems when trying to get 6 teams together to practice in the off season. Today, we use wifi which is more stable, but brings with it a different set of issues.

Now, what features would you be willing to give up for a new control system, and what features would you want added? Keep in mind that any feature you don't use, other teams undoubtedly will.

Now, for Support:

Having adequate support is a huge issue when we're talking about a control system as complex as we need. Realistically, you need a large company and millions of users out there in order to get the support to ensure no one has any serious issues. As it is, we get reps from NI at the regionals (there were two in Duluth) helping teams out, tracking field performance and fixing any issues that come up. We have an amazing support library for 3 languages that makes programming the robot dead simple (especially when compared to the old IFI system).


If anyone has a better control system, I would strongly encourage them to bring it forward. If you find something cheaper, more capable and with better support, I know FIRST will want to hear about it.

[AdamHeard]

Originally Posted by Jon Stratis

Up in Duluth two weeks ago comm issues were pretty rare. There were a couple of teams that had recurring issues (and we worked with them to try to resolve those issues, and in some cases succeeded), but there were many more teams that went out and ran every match without a single problem.

There are really two issues with having a control system: capability and support.

First, capability:

The old IFI systems were great, but they lacked in capability compared to what we have today. Back in Rack 'N Roll we tried to get vision tracking to work, and had a horrible time with the camera that was available then. With the new system, vision tracking is actually pretty simple... but it brings with it problems some teams encounter, like having too high of a framerate and bogging down the CPU.

Back then, you had to use Spikes to control all of your 12V solenoids. Today, you have the additional option of using 24V solenoids, AND don't need to control them with spikes.

Back then, power management was interesting. You had multiple distribution boards for high and low current applications, you had DIN rails feeding those, there was no visual feedback when something was wrong, and there were no regulated outputs. Today, the power distribution board has all of that and makes everything dealing with power much simpler.

Back then, you only had Victors. I know, a lot of people still prefer the Victors, and that's fine. But the Jaguars bring a whole host of increased capability, and I know our team has found ways to use all of that capability over the years.

Back then, you could only use serial connectors for your driver station unless you bought a special converter to allow you to use USB.

Back then, you had channel overlap problems when trying to get 6 teams together to practice in the off season. Today, we use wifi which is more stable, but brings with it a different set of issues.

Now, what features would you be willing to give up for a new control system, and what features would you want added? Keep in mind that any feature you don't use, other teams undoubtedly will.

Now, for Support:

Having adequate support is a huge issue when we're talking about a control system as complex as we need. Realistically, you need a large company and millions of users out there in order to get the support to ensure no one has any serious issues. As it is, we get reps from NI at the regionals (there were two in Duluth) helping teams out, tracking field performance and fixing any issues that come up. We have an amazing support library for 3 languages that makes programming the robot dead simple (especially when compared to the old IFI system).


If anyone has a better control system, I would strongly encourage them to bring it forward. If you find something cheaper, more capable and with better support, I know FIRST will want to hear about it.

When you compare to the next Vex controller, half these points drop away.

[lynca]

Originally Posted by Jon Stratis

If anyone has a better control system, I would strongly encourage them to bring it forward. If you find something cheaper, more capable and with better support, I know FIRST will want to hear about it.

A detailed comparison between existing robot control methods would be fantastic !
Here are a few potential platforms,

1. cRio: https://decibel.ni.com/content/docs/DOC-19103
2. VEXpro Arm9 : http://www.vexrobotics.com/217-2180.html
3. VEX Cortex : http://www.vexrobotics.com/276-2194.html
4. Open Source : Embedded Microcontroller with Linux Laptop

Does anyone know of other potential candidates for a robot control system ?

[akoscielski3]

well lets see. At smoky mountains our first quarter finals match 1 robot was dead on the field (3259). In first finals match our robot took about five minutes to connect, and our driver station had to be restarted, and then our vision tracking wasnt working the whole match. Finals match #2 234 died on the field in the first 30 seconds of the match, but barely squeaked out a win with a double balance with our other alliance partner. final score was 42-22. Even though with all these failures we didnt lose any of them.