What grade level was Vex AI designed for? Is it a high school competition or a college competition?
Both High school and college.
It is struggling to get off the ground in College though. Currently 5 college AI teams signed up.
How many hs teams signed up?
Looks like 45.
And I just wanted to add on the college front, there is not a single college signed up. Only private teams. No educational institutions. Sadly all serious big name University teams that were really excited by VAIC ended up losing interest when RECF laid down their ultimatum.
Which ultimatum was this?
It is all quite complicated. And you can read the thread here if you want to see more information.
But a lot of it comes down to RECF/VEX investing a lot of money into building a new “AI system”. The universities teams I would like to see in the AI program are all serious schools and full of students that want to learn everything they can. They are taking college courses that teach stuff like computer vision, AI, and machine learning; they want to use the knowledge they are learning. People like that don’t really want the VEX AI system. So the RECF has spent a ton of money buying hardware and developing software that these teams don’t really have huge interest in. (which isn’t to say no one wants them)
So what happened is RECF wants 2500$ for registration + hardware + software + support. Lots of teams want a way to just register, like all the other RECF programs.
RECF believes that given the money they have already spent building up the system it wouldn’t be in their best interest to let teams just register. Otherwise they believe they would have no chance of recouping the costs they spent building the system. (read up on sunk cost fallacy if you are curious)
There is a belief among the teams that a lot more teams paying 500$ in registration might help the RECF offset costs more than less teams paying 2500$. Without estimates of number of teams that is clearly impossible to know.
The ultimatum is really in response to attempts to negotiate some middle ground. The official RECF response is “take our initial offer or leave”
Oh right, the whole black-box-in-a-programming-competition debate.
when i first heard that analogy at the start of the discussion i immediately lost all interest in vex ai ngl
if im going to go into an educational ai competition to make an ai i want to actually learn how to make an ai
I don’t have a dog in this fight, other than liking the marketing video Vex put out on AI (and professional interest in AI generally).
That said, I don’t think RECF was terribly interested in a large number of teams for the first year. Especially given that these teams will not be able to re-use this year’s components next year.
I can understand wanting to keep a pilot program like this small. I can also understand wanting a small pilot program to have some participants that would be more representative of a target “median” ability level.
That said, restricting the pilot in such a way that elite ability level teams don’t participate does seem to be a short-coming.
One can certainly take issue with RECF stating that they are trying to recoup R&D dollars over the pilot program period, rather than a 2 or 3 year window. But these are covid-times, and money being tight may have forced this upon them. One would hope that in normal times, the “pilot program year” would be accounted for as R&D…
I don’t know much about VEXU, especially the distinction you make between “private teams” and “serious big Name University” teams are. I don’t know what University budgets are these days, and I could be off-base in thinking that $500 versus $2500 is a material difference to a Carnegie Melon or MIT (which I would regard as “serious big Name Universities” in robotics just to name 2). I’m also not sure that CMU or MIT participate in VEXU. My personal awareness of VEXU teams that would qualify as “big Name Univeristy” would be USC and Purdue (not throwing shade towards any I missed - that would be my ignorance rather than my judgement).
Come on Griffin - you took one sentence from my response. This was the entire response for full context:
The RECF has had tremendous interest from private colleges, 2 year community colleges and 4 year colleges. We made a decision to not purchase more equipment and keep the pilot at around 50 teams based on the feedback (including from this forum) that schools were concerned that they would not be able use some of the equipment after the first year. We decided a smaller pilot program is the better way to access where we are with the hardware and software and address some of the concerns brought up to us. As I said in my full response previously, we will learn and make adjustments.
I think other than a bit of snark, like comments about sunk cost fallacy, I did sum up your argument pretty fairly. I left out a few examples you gave for brevity, IT costs etc. People could follow the link I posted for a full breakdown.
I guess here I could have said “building and purchasing components for a new AI system” to clarify the cost includes hardware.
I made it very clear that the teams I am talking about that feel disheartened by the value proposition are not the only teams.
Honestly this I feel could have been your exact words. You explained the long term costs associated with creating the program and a breakdown of how the RECF needed to offset the large software development costs, and management costs. Fixed costs that don’t scale up with number of participants and benefit from being shared across many teams.
1 thing I found very important when writing the post was to make it explicit which comments were factual and which comments were beliefs. You notice I mention a belief among teams as well.
And I don’t attempt to insult the RECF by pretending to know for certain how the economic factors would work out. Both versions (RECF, and teams) are beliefs and are clearly stated as such.
Hi, I represent a team at the University of Michigan. VEX has a similar cost among the university programs. It is not a different cost to do run and operate a team at a community college or a major research institution. We are planning for this season to do autonomous robots, outside of the scope of VAIC, because we do believe it is a worthwhile challenge and quite applicable for many of our members to gain this experience. The materials provided by the VAIC registration would not be applicable to us, as we already planned to use separate hardware and in-house developed software. We have a strong interest in creating our own software for what we do, in order to give a strong industry-applicable experience that some people desire, we’ve also been in the process of writing papers to have our members be recognized in scientific journals for their efforts. We, as a team, see the cost of $3475 as quite high. As we would be utilizing the field modifications, the season registration, and the worlds registration; but not the provided hardware or software. The cost of VEX isn’t wildly different across VEXU organizations, the $3475 is equivalent to around 7 V5 kits or approximately 10 times our expected sheet metal cost. The person writing this post is the one in charge of the organization’s spending. We do not get access to the university funding unrestricted, we need to apply and gain funding from many sources(some outside of the university) to have enough for our yearly operations. Despite having one of the highest yearly budget allocations for VEXU, the VAIC registration would be a major percentage of our budget that is not affordable for our team to spend, and also construct robots. However, just because VAIC is not a reasonable pathway for us, doesn’t mean we won’t attempt the challenge of full autonomy. We would have loved to use VAIC as a competitive atmosphere, but unfortunately it is not feasible.
Thanks for the reply. I’m not sure I fully grok the issue. You state that:
Given that VEX (and it may be different for VEXU, pardon my ignorance) operates largely a “build-and-compete-with-a-robot-using-parts-provided-by-a-single-vendor” (especially on the hardware side), so this seems like it was unlikely to be a good fit.
I know Michigan is a strong FRC region and that ethos and stack would seem to align much cleaner to your hardware and software decision. Granted that FRC budgets are an order of magnitude greater than VEX. I know the lines with FRC between work “mentors” can do versus the “high school team members” is much, um, less pronounced than what I observe in VEX.
What drew you to VEXU/VEXAI? Is it the fact that an entity already has the infrastructure in place to come up with a reasonable game with relatively widespread adoption?
Often in business one is faced with a “buy-or-build” decision. It seems you may face similar here - you can “buy” into the VEXAI infrastructure (and restrictions), or you could “build” your own competition format. Frequently what is available to “buy” does not fully match the feature-set your users require, and one is left to decide whether it is worth it or not.
To use another analogy, it’s like you want to race F1 because you like the tracks, the point system, etc. but don’t like the restrictions F1 places on the car.
While little solace now, perhaps the VEXAI value-proposition will change next year when it is passed the pilot period.
i may have simply missed your point, but the idea behind vexai (and vexu) is to give competitors the freedom to machine and fabricate their own parts and software, without the restrictions that are imposed on vrc and vex iq. forcing those of us who want to compete to buy a pre built ai program and hardware that we don’t even intend to use is not really in that spirit
i do see recf’s side though, and i can respect their decision to do things this way, but it’s just not a good fit for me or others who were turned off by this
Our team is comprised of a large percentage of FRC alumni. The issue of mentor involvement in FRC/VRC is one that I will not comment on, as there have been shows of student-led and egregiously violation of the rules in both leagues. VEXU has a different ruleset entirely, we were drawn primarily by the ability to utilize manufacturing techniques and custom electronic hardware within a competitive environment, and a relatively low price tag associated with. We are interested in the VAIC competition format, but like you said in the Formula analogy, we are instead just making our own car to race on a track by ourselves for our own challenge. Competition is not the primary reason we build robots, but it is a fun aspect that we enjoy.
I respect and understand the RECF’s decisions and processes regarding the VAIC event. I understand that teams with my resources and goals are in a very small minority. While I would love a fully open ruleset league, it is horrible for the future of teams that do not possess the resources that we do. We are open to the idea of talking and assisting the RECF in their operations this season, but we will be conducting our business regardless how they decide what VAIC looks like.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.