NEW 2009 Control System Released

[comphappy]

I feel like people keep missing this:

What is the cost of cRIO? Will this change raise the entry cost for teams?
A No. Due to the investment being made by NI ($10M over 5 years) and NI’s suppliers, FIRST will
provide teams with a more powerful control system without raising the cost to compete. Because
there are many factors that determine the final price of the kit of parts, the exact entry price has
not been determined yet.

While it may be expensive hardware, they are making it available with out raising the costs of entry.

I just unwrapped labview and installed it from this years KOP, looks interesting. As a lead programmer, founder of our team, and soon to be senior, I must look out for the future of the team. While i am an ardent supporter and advocate of ASM and C, LabView has a lot of potential for sustainability, while holding the real world experience that EasyC and what ever that other thing was lack.

As for the all around control system it has a lot of potential, I am peeved that the FPGA will be very limited if not completely lock away from us, I think that would have left us all where we wanted to be with the full spectrum of options.

I am just waiting for the day when our robots have full server racks crunching numbers so hard we dont have to worry about who is going to drive the robot, unfortunately i don't think I will be a team programmer that day, but I sure hope I will be the mentor watching this realized.

[Vikesrock]

Originally Posted by comphappy

I feel like people keep missing this:

What is the cost of cRIO? Will this change raise the entry cost for teams?
A No. Due to the investment being made by NI ($10M over 5 years) and NI’s suppliers, FIRST will
provide teams with a more powerful control system without raising the cost to compete. Because
there are many factors that determine the final price of the kit of parts, the exact entry price has
not been determined yet.

While it may be expensive hardware, they are making it available with out raising the costs of entry.

Yes, they are giving us one without raising the cost of entry. However, to keep previous robots functioning a veteran team will have to purchase a new control system every year as FIRST will only give us the first one.

[comphappy]

I would not consider that to be a major concern, and based on how they bolded the word HEAVILY my guess is it wont run more then 1k, and how many of them do you need at once. I would figure one for the current robot and one for prototyping with the old one would be enough, the modules are easy to swap. I will reserve my final call to Fall/Early Winter when all this is finalized.

[comphappy]

FYI the announcement video is now up so my guess is the others are soon to follow:
http://media.wpi.edu/News/Events/Rob...nouncement.wmv

[Madison]

Originally Posted by comphappy

I would not consider that to be a major concern, and based on how they bolded the word HEAVILY my guess is it wont run more then 1k, and how many of them do you need at once. I would figure one for the current robot and one for prototyping with the old one would be enough, the modules are easy to swap. I will reserve my final call to Fall/Early Winter when all this is finalized.

We usually have four or five robots running at any given time. Keeping that up with the new control system -- even if it's heavily discounted to $1000 -- represents a significant increase in cost to our team. We may gain more processing power, but it will significantly affect how teams are able to reach out to their community to generate interest in FIRST, or to help new teams get their feet wet in the preseason.

Registration fees were raised $1000 to cover costs associated with providing each team a new control system each season. If continuing to provide a new system to teams each is no longer possible, the registration fee should be reasonably lowered again.

[Drwurm]

Our team practically depends on old robots for all of our PR. We try to get at least one to every demonstration. Not getting a new RC every year is going to hinder that.

I can understand that the new system may be too expensive to provide a new one every year; but I cannot understand why FIRST would have gone with such a pricey system to begin with. Sure, the OI/RC setup needed a little dust-off, but this is overkill. I don't think any FIRST team is near the level that requires this much power.

So far I am not enthusiastic about this new system, however, I will hold final judgment until I can get my hands on the final version.

[JesseK]

I'm enthusiastic about the possibilities of this system. There are huge ramifications of being allowed a laptop at the driver's station. Even if much of the communication between the laptop and driver's station is limited, we'll still have the flexibility for more driver control. This could be anything from advanced GUIs to GAs to sensor matrix analysis via Matlab standalone scripts.

For demonstrations, let's consider a couple of things:
1.) You'll still have your IFI bots to take to demonstrations
2.) It is the cRIO that we get one of. We should get a new PDB & sidecar every year. Since the cRIO itself is modular and takes less than a second to reprogram, it's possible to move the cRIO from bot to bot and reprogram it during a demonstration. Sure, we can't show cRIO-based bot-to-bot interactions, but we will still have IFI-based demo bots for the next couple of years.
3.) Our team only has 1 demo each year in which we take more than 1 bot. Every other time more than 1 would be too much. Is this not the case with most teams?

[thefro526]

I saw a little bit of the new control system when I was in ATL. It looks like the possibilities are endless for those who wish to push the envelope, but for teams like mine we won't really have a need for all of this extra power and features. Some of the more basic ones, like USB interfaces on the OI, are nice and easy for all of us to use, but then there are other more advanced features that are going to be useless to a lot of teams.

In my opinion the new system is going to be a nice change if introduced into FRC right. I personally think that next year's game should be one similar in game play to the last 4. I.E. simple-ish auto mode, tele-op goals that do not require amazing control and programing skills, and a way for teams to score with little to no effort. This game would be a good way for teams to learn the new control system and good way for FIRST and NI to work all of the bugs out.

Then in '10 you begin to push teams into using the actual power of the new system. Make it more advantageous for them to use some of the more advanced features now that they have learned the basics. Even still make it so that teams who don't have the resources to use the new system to it's full potential have a way to remain competitive.

My prior statements may be totally off base though, they are coming from the guy who designs robots and thinks three limit switches is a lot. j/k

[Mark McLeod]

I like the technical capabilities of the new system and I'm sure we'll be able to work around the limitations it places on us all, but it will take some adjusting to. The cRIO will open up more possibilities for inspiration. We love new challenges. I'm not worried that we'll see magic next season, but do like trying to foresee potential issues that we can prepare for now. The more information we can gather now, the easier it will be to advise teams.

I'm concerned about how turnkey the system will be for teams with few or no technical mentors. The multitude of cross wiring that all looks alike in this new system (multiple 5v, 12v, 24v power connections) worries me. We've all seen lots of examples of cross-wiring and loose connections by team's with little electronics experience that threaten to ruin an inexperienced team's experience.

Electronics layouts will take on a 3 dimensional design. While the cRIO is resistant to abuse, the connections to it are not. It makes me nervious to see those delicate convertors sticking up into the air off the cRIO brick. The brick will survive, but we'll need spares of the exposed connectors and analog and solenoid circuit boards. I imagine bricks will often be mounted on their side to provide some protection.

I am concerned about the development software licensing issues. I really liked the team-wide mcc18 license that allowed multiple students to learn to program at once, in the lab and at home. gcc will still allow this, however, Labview will not. I had similar issues with the limitations on the single team EasyC license.
I can't count how many problems I had with installing software on school provided laptops without an administrator. I'd meet the students at night and be leaving fix-it notes for the IT department in the morning. For example, EasyC stuck on Vex, because it had to write to a protected folder and then be restarted (not to pick on EasyC or school required IT protective measures).

Originally Posted by JesseK

3.) Our team only has 1 demo each year in which we take more than 1 bot. Every other time more than 1 would be too much. Is this not the case with most teams?

We, like Madison, also have five or more older robots that are used quite a bit. We typically loan robots out to pre-rookie teams for competing in pre-season events, to demo at their school, or just to gain experienced in taking them apart and putting them back together.
On a given day we've had one robot driving in the Homecoming parade, another couple competing at a local off-season event with pre-rookies, a robot off being presented to a school board that lacks an FRC team.
During build season development the older robots are used for prototyping concepts, driving against the new robot, etc.

If we (I really) can afford to purchase an extra cRIO, then we need it as a portable workshop to take on the road to teams and to drive prototyping platforms. Our future retired cRIO robots will likely have to be retrofitted with older controllers much as we do with our old pBasic controllers. Advanced capability will be lost unless we devise our own replacement system. I am concerned about any necessary upgrades in later years that may render an expensive investment in a spare cRIO worthless.

The expected exorbitant cost of the cRIO for subsequent years is not justifyable as an expense for any of the teams I work with, on a post-season robot. Many teams I deal with are on a very thin shoestring and must recycle parts off older robots in order to compete anyway, so this won't really be an issue for those really low-end teams.

[yoyodyne]

Originally Posted by Mark McLeod

I am concerned about the development software licensing issues. I really liked the team-wide mcc18 license that allowed multiple students to learn to program at once, in the lab and at home. gcc will still allow this, however, Labview will not. I had similar issues with the limitations on the single team EasyC license.

Wow! I kind of figured that along with the new WPILib being open source there also would not be per-seat licensing for the new control system development environment. I guess it was naive of me but I figured that is why we would be getting a "custom" NI build. If there is, it is really going to negatively impact the way our team develops and unit tests software. Do you think that there will be license costs from both Wind River and NI? If that is the case, then there will be an impact to the teams that take the C route as well.

[Joe Ross]

Originally Posted by Mark McLeod

I am concerned about the development software licensing issues. I really liked the team-wide mcc18 license that allowed multiple students to learn to program at once, in the lab and at home. gcc will still allow this, however, Labview will not. I had similar issues with the limitations on the single team EasyC license.
I can't count how many problems I had with installing software on school provided laptops without an administrator. I'd meet the students at night and be leaving fix-it notes for the IT department in the morning. For example, EasyC stuck on Vex, because it had to write to a protected folder and then be restarted (not to pick on EasyC or school required IT protective measures).

The Labview that is provided in the kit the past few years has been an unlimited license for FIRST use. See the following post: http://www.chiefdelphi.com/forums/sh...ad.php?t=51405

[Eldarion]

Originally Posted by yoyodyne

Wow! I kind of figured that along with the new WPILib being open source there also would not be per-seat licensing for the new control system development environment. I guess it was naive of me but I figured that is why we would be getting a "custom" NI build. If there is, it is really going to negatively impact the way our team develops and unit tests software. Do you think that there will be license costs from both Wind River and NI? If that is the case, then there will be an impact to the teams that take the C route as well.

I agree 100% with the above.

One other issue that I haven't seen brought up here is platform lock-in. With the old system, teams could use Windows, Linux, or Mac to program the robot with, and the robot code itself was fairly portable between microcontrollers. In addition, low-level C programming is pretty much standard in the embedded systems world, so students are gaining real-world experience with direct application to industry.

With the new control system, you are locked to Windows, regardless of the method you use to program. LabView will run on Linux; the software that it uses internally to generate the VxWorks image will not (BTW I don't think Wine will solve this problem unless you can get an entire LabView install into Wine along with the other tools.)

If you decide to use LabView to program, the problem is even worse. LabView programs cannot be used on anything that is not directly supported by LabView, or converted to another, more standard programming language! Also, I question the applicability of LabView programming knowledge in industry--sure, some large companies and colleges have access to LabView, but most do not due to the extremely high cost involved.

The OS lock-in problem will only get worse when Vista takes over--I have spoken with many developers who cannot stand Vista and have switched to Linux, finding it better suits their needs.

Just some futher thoughts, feel free to comment on them...

The new cRIO system looks like the cRIO-9012 model (specs wise) with the integrated controller and chassis. The price range looks like it will come to ~$2000-$3000 dollars for the whole system.

When at one of the presentations, one of the NI reps said that the 802.11a standard would be used. This would make sense because of the 11 channel system that it has.

The fact they are letting us compile with C++ opens up lots of oppurtunities. You could actually code in a different languages (provided that you could convert the C++ libraries to a different language) and then convert to C++ to use with the cRIO system.

I think that this new system is way better than the old. The IFI system always seemed very limited to me. You had limited memory. The system couldn't support higher level language and as a result class files/namespaces/polymorphism and other higher level language features could not be used. I'm really excited to get my hands on this new system. I heard that this system will be available for us in November, can anyone confirm this? Meanwhile, I'll practice LabView with mindstorms

[Adam Y.]

Originally Posted by Eldarion

With the old system, teams could use Windows, Linux, or Mac to program the robot with, and the robot code itself was fairly portable between microcontrollers. In addition, low-level C programming is pretty much standard in the embedded systems world, so students are gaining real-world experience with direct application to industry.

We are building robots not a microwave. In this case you need to change the requirments from a procedural to an object orientated language if you want to do anything useful. I've bought a few different books on robotics research and it always ended up having it skip out of procedural designs within two or three chapters.

[Doug Leppard]

Help me with this. As presented I thought this was a FLL-FTC-FRC solution. How does new system work with FLL and FTC? Do they have smaller cheaper systems that can migrate from FLL to FRC and beyond?

Or did I get this wrong and this is only for FRC.