How do you like deep space?

How do people who bought all that plywood feel? I fell like it is really unnecessary for 90% of teams, but they made it sound like people needed to spend $500 on wood before they saw the unveil.

I would change that a little bit. I would say “teams who prioritize and dedicate themselves to creating effective autonomous routines will do better”. I personally agree with my understanding of Jared’s statement that trying to drive with a camera would honestly be more effective for those teams who would otherwise throw something basic together for autonomous. And that’s unfortunate because it means those students aren’t learning the important lessons and skills that comes from autonomous design and programming. Better incentives and resources for programming good autonomous routines would be more inspiring and effective in terms of FIRST’s goals than is handicapping those who can’t find the time or have the skills necessary to do such with the old style of gameplay.

TLDR: Just making an autonomous routine that drives forward ain’t going to make any robot particularly successful.

5 Likes

We didn’t buy plywood until today. Even then we didn’t buy until seeing what we wanted to build immediately; more will be coming soon.

When I left the shop we had a Level 2 mostly done and a Rocket frame being laid out.

Matches will probably be interesting to watch, despite the slower pace. I could see more matches than usual ending in ties though, with the way scoring works. I could even see Worlds (or district championships) ending with both alliances hitting the scoring cap and it becomes an endurance test for the robots and drivers.

I’m disspointed no field game componenents were availabe to physically measure at launch and to help visualize scale

As for the game , we’ll see

1 Like

They could have one volunteer for each team who’s only job is to cover the eyes of the driver :slight_smile:

2 Likes

The sandstorm period will still be referred to as “autonomous” within the software, and will still follow the same pattern as before. i.e. there will still be calls to “AutonomousInit” and “AutonomousPeriodic” during the first 15 seconds of the game.

(And I can’t remember where I read that. I couldn’t find it in the manual when I went to look for it, so it must have been somewhere in screensteps, but I know I read that some time on Saturday.)

2 Likes

It’s okay, I guess. I just feel there was LOADS of lost potential in Deep Space. I would make a few changes to spice things up:

  1. Keep traditional auto, but make it so that the sandstorm will happen for 10-15 seconds in the teleop, preferably at random. This could add to the “Space is crazy & unpredictable” thing.
    a. If you’re going to keep the current sandstorm, they should at least have some form of bonus for adding hatches & cargo in sandstorm.
  2. This is out of left field, but maybe you could have a mechanic where there are “fuel tanks” full of liquid that could be moved by robots from the tanks to a hole in the rocket and maybe filling it up could net you a knuckle & foot’s worth of points or an RP.
  3. As far as I can tell, there’s NO reason to go high on the rockets in Playoff Matches other than more points… They should give bonus points for that.
1 Like

Yup! It’s in the Field Management System White Paper. The wireless network bandwidth is limited to 4Mbit/s this year.
https://wpilib.screenstepslive.com/s/currentCS/m/troubleshooting/l/705152-fms-whitepaper

1 Like

I believe that teams that have had a strong autonomous mode in recent years will put a lot of their energy into semi-autonomous features like robot alignment at compartment or loading station; game piece pick-up and delivery and HAB platform climbing.
All those features will have for objective to increase the performance and reliability of game actions and reduce cycle times.

1 Like

The lack of robot interaction is very disappointing to me. It’s not as bad as Recycle Rush, but it approaches that bad. It much better for everyone to have robot paths crossing.

2 Likes

I’m here to preemptively apologize to every FTA and CSA who has to deal with this and explain to people what’s going on. In 2017 it was field reset who took the brunt of the game…this year it will be the technical volunteers.

1 Like

I think that many of the top teams who do have traditionally strong autonomous modes will continue to do so, but also write their code and routines in such a way that it doesn’t take a ton of reprogramming to also apply that same code to semi-autonomous routines during teleop like you are talking about. Having a full autonomous that is consistent and point heavy will almost always be more effective and efficient than camera driving, just because you can fine tune it a lot easier without the human variable. However, I’m sure there will be some teams with traditionally camera-heavy drivers who will probably just devote most if not all of their autonomous time to semi-autonomous routines instead of the other way around.

3 Likes

I was really hoping for tennis balls, but bean bags would have been a very interesting concept.

1 Like

That’s true, it has to be a good auto routine

I think most teams try to have good auto routines, and then a lot of them end up driving in circles anyways. Those teams, especially when their drivers have traditionally used alignment cameras in the past, will probably ignore traditional autonomous and just add semi-autonomous routines in place to assist their drivers with the skill they already have.

Which brings out: what will the ultimate autonomous achievement be in Deep Space?
I suspect Hatch Panels will be the thing, but how many? 2, 3, 4… or more Hatch Panels in auto?
My bet will be that the top tier teams will ask their alliance partners to keep the path clear to a rocket and they will go for the top 2 rocket stages. I think Cargo behaviour in the Depot or out of the Reload Station will be less predictable than Hatch Panel delivery at those stations. And Hatch Panels have to be installed on the rocket before cargo can be loaded anyway.

1 Like

I think that this year’s autonomous capabilities are honestly a lot more complex than previous years. Instead of having a few specific ways of earning points, teams can earn points during auto by placing hatch panels and cargo in many, many different physical locations. Thus, I’d say that the most useful autonomous for a larger team would probably be a modular routine, where quals / playoffs alliances could strategize scoring before starting the match and have that data inputted into autonomous. For example, team A could place hatch panels on certain hatches and then team B could follow that up with Cargo and vice versa. With lower-caliber team’s ability to drive by vision during auto, this strategy becomes a lot more feasible too, especially if said lower-caliber teams have talented drivers.

Also note that the only hatch panels available to teams during autonomous / sandstorm are the hatch panels preloaded in the robots and the two hatch panels preloaded in the loading stations (since human players can’t load more during this period). Thus, without complex setups between alliance members, the theoretical maximum number of hatch panels for a given high-caliber team is three.

1 Like

I think 3 hatch panels in the first 15-18 seconds will happen. A team like Stryke Force with that speedy swerve drive will preload a panel and place very quickly on the front. Then go grab panel 2 and place on the other front. Since those are easily visualized from any driver station you don’t need any automation to place them with accuracy.
Then go grab 3rd staged panel and go to the nearest bay on the cargo ship.
Remember it doesn’t have to be “placed” in sandstorm it just has to be so close to the hatch that it keeps the ball from exiting. If you are good then you are done. If you miss a little then you have a few seconds to reposition the robot while keeping the cargo in the ship so the panel sticks.
I anticipate high level teams will score the 21 points in the first 15-18 seconds.

1 Like