Very exciting updates from the FRC Blog including AprilTag family being changed to 36h11 for 2024, new radio testing!
Congrats @kiettyyyy and Vivid Hosting for helping us get to the next level of radios for FRC! I’m very much looking forward to a radio designed for FRC by people who actually DO FRC!
Past - Usage Reporting
I have some exciting technology updates to share for this coming season, next season and beyond. But before we get to that, I want to share the usage reporting data from this past season. The FIRST ® Robotics Competition software has a feature called ‘usage reporting’ that tracks what WPILib objects are created in each team’s code and reports that data back to the field when the robot is connected. Data from the 2023 season can be found here. Some notes about the data:
The data has been anonymized. Team numbers have been removed and the data has been re-sorted so teams are not in order by team number. We can only track the objects teams create in code. If a team creates extra motor controllers that aren’t on the robot, they will still be captured by this system. If a team creates motor controller objects of the wrong type, that wrong type will be captured by this system.
Some objects naturally result in double counting (e.g. Encoders use Digital Inputs).
Counted objects and TRUE/FALSE show the largest number of any given object used in any one match (i.e. if a Robot plays match 1 with 3 Encoders and match 7 with 2 Encoders, 3 will be reported). This means that if a team switched motor controller types, or IMUs, or anything else during the season, the sheet will show numbers for both devices. Language and Framework report what was used in the last recorded match.
Some bugs in the reporting resulted in Phoenix Pro devices and WPILib PS4 controller data being recorded incorrectly. This information has been removed from the spreadsheet to avoid confusion.
Present - AprilTag Update
While the general feedback on the 2023 implementation of AprilTags was pretty positive, we did learn some lessons that we plan to apply to 2024 and beyond.
The first lesson was regarding tag family. We chose the 16h5 family for the 2023 season as our testing showed increased detection range for a given resolution. In practice teams were often finding the data too noisy to take advantage of the additional range. Given that, for the 2024 season we will be moving to the 36h11 tag family. This will substantially reduce false positives with default settings and is compatible with a wider range of software implementations.
Many of the other lessons learned were minor items with the physical implementation of the tag plates. The plates used for 2024 will be revised based on these lessons but we currently expect they will be compatible with 2023 plates such that teams could potentially sticker over their existing 16h5 tags. Some 2024 field tags will be mounted differently, but we expect team versions will likely be able to use similar tag plates. Tag size will be tweaked slightly to 8.125 in overall, with a 6.5 in. internal black square.
Future – Radio Testing
Over the past year, FIRST ® has been working with Vivid Hosting on designing a custom radio for FIRST Robotics Competition teams for use starting in the 2025 season (the 2024 season will continue with OM5P radios). Vivid Hosting may be new as a supplier to FIRST, but many of their employees have extensive team and volunteer experience and we are excited to work with them on this device. Features such as power connectors, device size/shape, and number of ethernet ports and their layout have been directly driven by their experience as team mentors and volunteers. We are planning on testing the prototypes of this radio at the following events:
Chezy Champs: San Jose, CA September 30 – October 1
Oregon Girls’ Generation & FIRST Chance: Wilsonville, OR October 7 - October 8
Battle of the Bay: Alton, NH November 4
Madtown Throwdown: Madera, CA November 11 – November 12
Here’s a quick rundown of features we think teams may be interested in:
Size: 4.5x2x1.25 inches
Mounting: Cable tie notches
Power: Passive Power-over-Ethernet (PoE) or wire-to-board connection. Designed to operate directly off battery voltage (though has also been tested downstream of a VRM or RPM)
Frequency: 2.4GHz + 6GHz dual band. At home, teams can use the device as a 2.4GHz AP to maintain compatibility with legacy devices. At events, the device can be used as a 6GHz bridge.
Performance: Among the many benefits, the 6GHz spectrum has more channels, allowing events increased flexibility in setting up on a clear channel. In addition, OFDMA (orthogonal frequency-division multiple access) has been integrated to make better use of wireless spectrum. These features, along with other 802.11ax improvements should contribute to a substantial reduction in matches with wireless congestion issues and will hopefully allow us to increase the team bandwidth cap in the future.
Beyond – Mobile Robot Controller RFP Process
We’re also already looking way out to the 2027 season and starting to think about a new Mobile Robot Controller after the roboRIO 2.0 agreement expires in 2026. 2027 may feel a long way off, but for a major project like this, it’s practically right around the corner! We have created a preliminary draft of a Request for Proposal (RFP) and will be sharing high level information with some existing sponsors over the next few weeks. We will also be collecting feedback from a small number of pre-selected CSAs, FTAs, LRIs and team mentors. After processing that feedback into the final document, we will be releasing the RFP publicly at the beginning of November.
This radio upgrade leaves a great opportunity to improve on the FMS, especially the very low bandwidth. I really like the increased number of Ethernet ports, it reduces the need for a network switch, which will decrease the number of failure points.
Yup they even specifically called this out in hopes of increasing the bandwidth. Very exciting possibilities!
“These features, along with other 802.11ax improvements should contribute to a substantial reduction in matches with wireless congestion issues and will hopefully allow us to increase the team bandwidth cap in the future.”
My only concern so far is additional mounting methods. @kiettyyyy would you mind explaining the rationale for that or is this open to changes in the future? Would the recommendation be to VHB or use extra strength velcro in addition to cable ties?
Finally a new radio, I hope this means a hotspot won’t cripple an entire field of robots. The 36h11 tag family is a great improvement meaning less tuning for false detections.
I really hope there is still a way to use the new radio as an AP outside of the 2.4GHz band, which is so crowded as to be unusable in many locations (including our shop). I would not want to have to swap radios for home vs. event use.
I’m also hoping that this is basically Wi-Fi 6/6E which would include 5GHz as well as the stated 6GHz band support. I’d be surprised if 5GHz isn’t supported by the chipset, even if not called out here. If it’s programmatically disabled, that would be very disappointing since most teams’ laptops won’t have Wi-Fi 6 support at this point.
These are huge, huge positive changes. The radio is hands-down one of the worst parts about competition, so a new purpose-built option really is a game changer. I can’t wait to see the results!
36h11 tags are also awesome. I didn’t expect the change, but I’m more than happy to start testing with 36h11 again this offseason. I expect big accuracy updates with this move.
Are these gigabit or hundred megabit? I hope we will be able to turn off the passive PoE outputs, preferable with a physical switch, so we do accidentally fry non-PoE devices. If not a switch, maybe default to off until enabled in code?
I hope we will be able to use the 6GHz band at home, and that there is also a 5GHz radio. We quickly run out of free airspace at out shop in the 2.4GHz band between the school’s APs and running multiple robots at once.
There’s a few options I could see. Hopefully it can be used as a 6 GHz AP, in which case 2 radios could be used point-to-point (using a radio or just a COTS USB adapter on the DS side too, if the DS computer is not 6 GHz capable). I’m not able to think of a technical reason why it couldn’t be run as an AP at 6 GHz as well, so I wonder if this is just not stated sufficiently clearly in the blog post writeup. If not, one workaround could be to buy a multi-band 6 GHz capable AP and have the robot and laptop both connect to it?
Idk about accuracy but in my very unscientific testing I found a substantial decrease in false positives with 36h11… by which I mean every shadow on my wall didn’t get picked up as a tag with that family.
But I do hope it means we get more focus on using tags for localization than simply for scoring alignment.
Yes, this would work as a workaround, as long as we’re able to run the radio as a client in that band outside of event use. (We often do this already with a 5GHz AP, actually.) It’s still a little clunky to have to bring an AP to run the robot at demos and such, but would be manageable.
Not supporting 5GHz and requiring new Wi-Fi hardware (AP/laptop adapters) would also not be great, but also could be worked around.
Another workaround could be to just swap in the old radio at home if you need it. You’ll also need to add a switch if you’re using more than one port, but the radio configuration tool should still support it for a while if the support for the D-Link is anything to go by.
This is all completely true. Limelights especially had a high rate of false positives so this should help.
I just noticed the note about Phoenix Pro messing up reporting. Hopefully this gets fixed for next year, as I’m sure a significant chunk of teams are on it, but now we don’t know exactly how many. They could at least have been reported as normal non-Pro devices.
Control system RFP is also a great thing. While I don’t know if the RoboRIO can be replaced with something cheaper, it would be nice to pay closer to $200 for something with less built-in ports and bring the cost of the system down. That also would allow non-FRC usage.
The new tag format! Thank you for listening to feedback from a lot of us out there using them.
Excited:
The new Radio!!! I can’t wait to go digital spelunking on one. Seriously, I hope I get a chance to play with one. It’s very exciting.
Worried:
Not just the possible lack of gigabit… which isn’t all that bad from the standpoint of DS<->Robot traffic but it can impact some of us that are more sensitive to latency when doing hefty file transfers off the robot. I’m hoping FIRST keep enabling us teams to use our own ethernet switches.
I’m actually far more worried about the POE ports killing devices for teams that don’t know better and hopefully they can be software disabled completely or even hardware jumper disabled. “Passive POE” is NOT a standard like 802.3af. It’s true to say that more things DON’T use it than DO and that plugging in raw DC power into ethernet lines is far more likely to release magic smoke than not. This is probably a great thing for the teams out there that have several limelights but a lot of other folks are just using Pis (both Raspberry/Orange/Green/etc). I’d like to see those ports defaulted to off and only enabled upon explicit request with some obvious LEDs.
It’s also possible, since this isn’t a standard, there is some safe way to do it but I don’t know what it would be.
I’m also very curious about the input power situation. REV has their POE thing spec’d for 18V and 1A max and the VRM does 12V and 2A max… so if the POE output is above that… that’s going to cook something possibly. So are the POE outputs only active when on DC power? Are we getting a new POE injector?
Loads of questions on this still but I’m sure we’ll get answers.
Curious:
What might be changing about the field side APs. If anyone going to the events can get pictures/details - I’ll buy you a cup of coffee.
What the software configuration for these new radios will be. What platform/stack it is using and what else it might be able to do to help teams with monitoring network communication.
Amused:
Fun fact, because we send the “Language” thing back to this inquiry as ‘900’, they do some decimal to binary conversion and that ends up truncating it to “1”, which is LabVIEW. That’s definitely us. Proud to continue that tradition.
EDIT:
Super Excited:
We will also be collecting feedback from a small number of pre-selected CSAs, FTAs, LRIs and team mentors. After processing that feedback into the final document, we will be releasing the RFP publicly at the beginning of November.
Seriously, can’t wait for this. The future is a fun place to be!
One thing I didn’t see mentioned on the new radio is the power up time. I hope that its faster than OM5P. When we have 2 minutes and 30 seconds of match time losing 60 seconds because your radio decided to reboot is a terrible feeling. Even better if the new radio doesn’t reboot randomly(wire to board connection should help).
