New Products - June 2021

That’s because of the okapilib which has built in odometry. Also it really depends on the robot. x-drives or mecanums are way easier to do motion algorithm stuff with than tank drives.
I’m personally not using the “gps” as most of the fields at local and state competitions don’t have the tiles and walls mesh perfectly.
Another thing to note is that the “gps” has worse accuracy but keeps a consistent accuracy while odometry once it gets off slightly it can become much worse than the “gps” in the long run.

1 Like

You have no idea how accurate the new GPS sensor is. We have made no claims as to the accuracy and this is not the GPS that’s been available for the AI competition to date.


This is really only the case for poor quality GPS. High quality stuff can get 1cm precision.

And I find no problem with VEX using the name GPS for their system, if you want to learn about the history of the name GPS see this post

It used to be called " “NAVSTAR” (Navigation Satellite Timing and Ranging)", a very specific name that says what it does. The current name GPS really just describes any system that provides robust drift free global position. ( Global in this sense not meaning on the globe, but not “relative” to some initial starting position). The VEX system is clearly a global positioning system, and given it is such a generic and well-known term it makes sense to use it for this product.

There were questions raised about the quality of it with vex fields flexing and being slightly different spacing, and those are fair criticisms but we will have to wait and see how big of a deal that ends up being.


Tabor for GPS


So this is a completely separate system from Vex AI. Does this mean that VRC and VEX U competitions should plan on having the field code strips at every field in the events? I don’t see how we could maintain consistency through competitions if it’s not a requirement.

If it’s for remote tournaments, then of course it makes sense to buy at your own discretion, but that probably won’t be the case for most people.


What I meant is that this is a new sensor, it’s similar but different (it has improvements) from the Flir camera/Jetson nano system that was offered for the Change Up AI competition.
See Bob’s comment here.

I have no insight as to what the GDC or RECF may recommend regarding installation of the field strips.


Yep, with about $1000 worth of hardware and Skylark’s pricing structure includes a monthly plan for $50 per device and an annual plan for $495 per device.

I’ve been a GPS user for a long time, so having a local signal source is nothing new.

I regret kicking this can of worms down the road. While we all want to hear about cool VEX products, this is one that needs a lot more documentation other than the CAD drawings of the case and ships in 6 weeks. I’m truly excited about getting better than 16 feet of location accuracy for a few hundred dollars.

Had it been released with links to the docs Bob talked about I would have spent today burning up credit cards with preorders.


Honestly although it is cool they added a GPS sensor…

  1. $200, are you serious? How come VEX hollers “No 3D printing components in VRC until everyone has access to a 3D printer!” for 10 years, then recently make a product that costs almost as much as the V5 Brain. By the way, nowadays students can buy a 3D printer with pretty good quality for the $200 price tag, maybe even less. Not to mention, there is such thing as ordering 3D components as well which allows students the opportunity to get 3D printed parts without having a 3D printer. Like… Let’s see the choices:
    a. We allow 3D prints to allow students to get used to CAD designing and encourage building designs in software.
    or b. We ban 3D printing designs for robots, and instead we create a product that costs about just as much as a V5 Brain, and it takes the aspect of learning and applying trigonometry in code away. Students no longer have to learn to innovate code in more advanced ways. Now they can solely rely on the GPS sensor to tell them that they are off-course and move the motors with a P control loop, which would probably be as competitive to those with sophisticated odometry and motion profiling.
  2. Now that pneumatics and 8 motors are the limit, every team will be having to spend an additional $150-200 for pneumatics. Yet again, where’s this ‘even playing field’ VEX was trying to provide, which is supposed to be their motive?

Am I mistaking something? Do I have the right to be mad, or am I missing the mark of what’s going on here?


Ok so you’d rather VEX not release a product if it’s not cheap? Where’s the sense in that? If anything I’d say the $200 price tag indicated that this sensor might be of some decent quality.


I guess the conclusion is that the 99.9% of people who compete in Vex (MS/HS/College) currently don’t know if this sensor will be of any use to them.


Based on implementation, purpose, and effect-it depends. If the product is expensive, it must have essential importance and impact. For example, the V5 Brain is $250 for one unit and has importance to allow the system to run. Everyone HAS to get a V5 Brain to compete, and the knowledge of “This is my ticket to compete, and I understand that paying the money allows me to learn more about robotics.” But what’s with the VEX GPS? Does it teach me to code odometry? Does it teach me about being more precise in autonomous? Does it teach me about learning methods to compensate wheel skid? Or does this sensor just give me the very things that I am supposed to be learning through the EDP, but instead VEX does my own homework for me if I were to spend $200. This is the same exact aspect with VEX pre-packaging PID into their EDR motors, so students who do not know PID are less inclined to learn PID because the homework is already done for them. It feels like cheating, except instead the teachers are giving out the answers to all of the students, then after tell the parents “look how successful these students are!” who did nothing but copied random jibberish on their papers.


GPS rewards teams for having funding, not for being good at designing a good odom system. In fact, it actively punished teams that have worked hard on odometry by making their work pointless.

My ideal version of vex? Nothing ks just given to you. All you have is some brushless motors with encoders and with direct voltage control. You can also use cameras for CV if you want, or any of the V4 sensors, but nothing is just handed to you. Making V5 “smart” is the worst thing to ever happen to the Vex robotics competition, because it gets rid of reason to learn more.

Oh and of course 3d printing because its 2021. I have a decently cheap Prusa Mk2S printing a good Benchy in under half an hour. If that printer can do that, anyone can afford a decent printer of they can even afford to compete. I do think unlimited pneumatics was the right call though, since they were p much useless before.


This sounds like a good way to make sure that as few people compete in VEX as possible.


may I remind you that v4 didnt really hand things to teams?


There will always be a tradeoff between approach-ability and “DIY”-ness. Clearly, VEX chose to go towards the approach-ability side of the scale with V5. Maybe you don’t agree with this decision, but it is still a valid one.


Oh so you want FRC? Cool, do that then. I’m not sure why you’re trying to get VEX to be something that it’s just fundamentally not.


the problem with Vex’s approach to raising the floor is that it does nothing to raise the ceiling. Now there is no innovation to be had, and not as much to work towards for the best teams. It would be different if say, the vision sensor allowed you to implement your own CV algorithm, but you cant do that because theres not enough oower or bandwidth to even think of such a thinf.

1 Like

So you go work towards the ceiling and let people closer to the floor be happy that they can do cooler stuff.


Hey Drew, go max programming skills, then come back and talk about how the ceiling isn’t high enough.


I want vex to be something it was, the greatest 2 minutes in robotics.