Wheels will never be usefull with the current terrain generation (especially on Verdure)

ToastyCosty shared this feedback 2 months ago
Under Consideration

With vs 2.3 i've had the pleasure of starting to build rovers, on of my favorite activities in SE1.

After at least 8 hours of testing/driving, while i am aware the current iteration of wheels still needs work, i am 90% sure wheels will never be able to even come close to the usefullness of their SE1 counterpart, and it's a problem mostly related to the terrain generation in SE2.

While SE1 had bumps, hills and mountains, SE2 completely re-defined how planets are generated by adding MUCH MUCH more terrain complexity (absolutely beautifull to look at, terrible to drive on), most of it being steep inclines, ridges, and so on, making navigation extremely complex (also more trees/rocks).

This makes wheels more suited for "smoother" terrain such as Kemik, reched much later in the gameplay.

Since wheels are a staple of early game building, and the player starts on verdure, i feel they may be a trap for new engineers that think building and driving a simple rover would be easier than going straight for a ship, but the reality it's the complete opposite.


Unless wheels get unrealistic amounts of grip and stability, they will always be worse than flying around with a ship and bypass the terrain.

Replies (19)

photo
6

I did a 'sort of' planentary survey (not comprehensive) and came to the conclusion that approximately 20% of planet verdure is driveable in any real sense by anything but small rovers that weave in between the trees, but which could not handle any real obstacles short of driving around them.

Of that 20% driveable terrain, 90% of that will be underwater in VS3.0, which makes it incredibly limited to run around in any sort of wheeled vehicle.

I like the rough appearance of Verdure, the tree cover and the like, it just makes rovers largely impractical. Slower, prone to flipping and blocked by medium sized rocks (plus some trees/plants seem to be invulnerable and imprenetrable), rovers are fun but not the best way to get around.

I haven't found it yet, but I was kind of hoping for an actual desert space that was flat-ish with dunes on Verdure so you could buggy like crazy. Even the beach spaces are packed with palm trees. Granted, I'm driving a giant rover (Think Grover) so trees are more of a slow down because of knocking them all over.

photo
2

I found a desert that is somewhat drivable: it's a long strip near one of the "future" lakes and it has a low-ish amount of trees on it, the dunes seem smaller than other parts of the planet and rovers are still prone to flipping over, but i feel that's the best we got for now.

here are the cords: 24280.96552423373 -2745.2345429650086 -298167.07338153006

photo
photo
2

I've been trying to drive, and I'm encountering a lot of "waves" in the terrain. Terrain that should be smooth from a geographic perspective is very bumpy, which makes driving a pain, as you are aware

photo
1

i was just about to start building a rover and i check here and see this... >.< im curious to see how long it takes me to travel the 40km to my target area.... wish me luck i guess?

photo
1

I have a rover, and it seems remarkably durable. My main problem is that sometimes one of my tires will clip through the mesh and *launch* me. The first time it happened, my vehicle and I exploded. The second time, we were fine. Of course, the second time it was much less dramatic

photo
1

I do recommend this: Have 8 (eight) gyros, split in 2 (two) groups. One (one) group will be "Primary Gyroscopes" and the other will be "Auxillary Gyroscopes". 8 (eight) gyroscopes is too many for most of the time, but sometimes you need 8 (eight). It's always handy to have backups, especially when you roll. I have them spready out across the hull of my vehicle, set into the armor

photo
photo
4

In addition to wheels physics enhancement, I strongly suggest the devs to balance resource costs and add road building mechanics to encourage ground engineering.

photo
4

Ooooh, a terrain flattener attachment would be cool!

photo
5

I sugguested such a thing this morning. Perhaps there will be a future 'Earthworks' or terrain modification tool to make road infrastructure.

photo
photo
8

Rovers and boats (when we have water) should give you a real benefit when transporting heavy loads compared to flyers. They might be useful for working with terrain as part of heavy-duty mining equipment. Those kinds of machinery should give you tangible gameplay benefits. Using flying mines on planets should be less practical, especially on a large scale. At the same time, not all resources might be found in space on asteroids, otherwise all the benefits of the rovers will be meaningless if you can find and mine everything in zero-g convenience or buy at the trade station. Rich deposits of rare resources can be found primarily on the surface of planets with harsh weather, where using rovers can be one of the best options. So yeah, rovers can and should be useful at any stage of the game, not necessarily in early game only.

photo
5

Hazardous weather on a planet could make Landing tricky at best, dangerous at worst. Drop in from space, crash, build rovers, and try to figure out how to Rocket back up into orbit with your loot.

photo
4

Honestly? Boats should have the highest energy efficient propulsion, followed by wheels, followed by thrusters. The real benefit of a boat should be the easy energy management. One or two solar panels to power a ~30m craft with decent manueverablity. A rover should probably need the energy of 2-4 solar panels for a similar ~30m craft, and air and space craft frankly should require 3-6 solar panels worth of power. But for aircraft obviously adding solar panels would increase the load and decrease the value of the craft without a base of operations.


Maybe a hydrogen generator could offset that for air/spacecraft to have steady power before you get to uranium. Higher resource cost for greater independant action.


But the problem is that the current wheel implimentation makes rovers more complex and less effective than an aircraft with a base and counter productive (at least on verdure).

photo
4

this is one of the reasons i was looking forward to wind physics. ive already dabbled in a "wind diven" land cart.

it was not super effective, but we dont have canvas yet, i cant make a proper sail, and wheels are still... getting inspiration from the community.


eventually if we can get air foils, i would very much like to build a tethered glider that sits in a wind current providing power from solar pannels. just because it would be fun.


nevermind eventual possible hydrofoils.

photo
4

oh one other thing... because terrain is as complex as it is, the "engineering solution" to make a vehicle that can pave roads, level pathways between key points, make the use of vehicles worth while by creating a system to make driving them more functional.

photo
photo
3

Reminds me of the vsauce video "Why don't any animals have wheels?". In a nutshell, wheels are useless without roads, especially on M-class planets covered in trees. Well before vehicles, humans built roads. I'm +1 for some better terrain manipulation tools, I'm -1 on just making it easier. Verdure's terrain could use a little massaging though to be fair.

photo
1

Almost like we need horses and mules for rough terrain...or tank treads.


Or, a 'mechanical horse'...?


Mech legs/Mech leg controller block!

lol


Someone posted that idea, and I just really like it, I hope it works out someday ;)

photo
1

Always planned on making a hexapod mech once programmable blocks came out, vibe code the fuck out of it. but a dedicated mech leg controller? sign me up

photo
1

No need to, a mod just came out on SE1 where you have different sizes/types of mechlegs, with full left/right pairs, that you just slap on the side of your grid. Building and using them is just like you do with wheels, and they just work. A little janky, but Keen could clean that up. SE already has the logic for 'sticky' feet, for when your character is walking on surfaces in zero-g, so borrow that so your mech isn't always tipping over. Enable a slow walking mode that uses a landing gear 'lock' to the terrain for walking up mountains, or upside down in gravity, and your mech can operate in rugged terrain.

photo
1

With water coming, there will be a set of water thrusters and boat/submarine parts, so your initial transportation types for survival would fall into boats for water biomes, mechs for mountains/trees, and wheels for moon buggies or desert type biomes. Then air thrusters for atmo planets, hydrogen/ion for non-atmo, and they need some sort of final space engine to unlock, either a nuclear plasma/NERVA/fusion engine type as end-game tech.

photo
photo
2

Im tempted to agree but until wheels have parity with SE1 i cant say for sure whether or not I totally agree.

That said id recommend the team take a look at adding natural pathways, dried rivers or rockslides forming smoother areas to drive up and around mountains or other harsh terrain; theres plenty of hand waving that can be done to explain natural 'roads'

photo
2

Perhaps it would be good to start on a smoother planet where wheels do have a usage, or yes some kind of erosion tool, something that allows you to flatten terrain or add some material that makes it flat but does not allow you to dig deep holes super easily.

photo
1

I do hope that by the time water is ready, Keen have taken a good look at Verdure and reworked its terrain to better serve gameplay needs.


It's beautiful as is, but the terrain variation is far too extreme for how small the planet actually is.

The mountains and oceans need to be more spread out with flatter terrain in between, or the planet needs to be increased in size significantly, or both. Trying to drive around at present feels like every surface on the planet is on an incline, which also makes finding a nice spot for a base inconvenient.

photo
2

The solution could actually be quite simple—the height of a mountain range should be proportional to the planet’s radius... On Earth, this ratio is around 1–1.25 per mille.

The maximum height of a mountain range is limited by the bearing capacity of the underlying strata, or, from another perspective, by gravity, gravitational acceleration, and the thickness of the planetary crust. If a mountain range exceeds a certain threshold, the pressure in the subsurface will exceed the subsurface’s load-bearing capacity, and the range will begin to sink into the planetary crust and mantle.


The area covered by mountain ranges is also vast—on Earth, areas higher than 1,000 meters above sea level make up about 1/3 to 2/5 of the land area, and areas higher than 2,000 meters above sea level make up only about 15% of the land area.


Counterargument—but on Mars, Olympus Mons is twenty-two kilometers high!

Yes, but!

1) Olympus Mons is actually very flat—it has a diameter of nearly 600 km... so its slope has an average gradient of less than 5°

2) Mars has lower gravitational acceleration—only 3.7 m/s²

3) The Martian crust is much thicker than Earth’s, estimated at 40 to 80 km, and in some southern highlands it can reach up to 100 km. On Earth, the planetary crust has an average thickness of 5 to 10 km beneath the oceans (about 2/3 of the surface) and 30 to 50 km beneath the continents (about 1/3 of the surface), with maximums of up to 80 km beneath major mountain ranges. Mars, however, has only about half the diameter of Earth, so its relative thickness is even significantly greater.


The area of mountain ranges on Mars—regions higher than 1,000 meters above “sea level”—comprises about one-third of Mars’s surface.

photo
7

Wheels as they are now have unrealistic low torque and grip. You see irl vehicles on 4 wheels going up very steep hills, while in SE2 they barely can go up a 30 degree slope.

photo
5

Except when braking, then they seem to grip so well that you flip :D

photo
4

The grip/friction is mostly fine, it's the default acceleration values and lack of wheel torque that produces the lackluster wheel performance.

In my experiments, the default acceleration values make rovers move like molasses. In order to make them accelerate like electric motors would, with instant torque delivery, you have to max out the acceleration value. Unfortunately, doing this also makes them brake equally aggressively, which results in front-flips when you apply the brake. If there weren't enough friction, the whole rover would just slide/skid a bit during aggressive braking, rather than flipping over forwards.

There's another thread where myself and other folks have outlined a number of changes that wheels need to be more viable, but a short version is basically that we need more configuration options (and probably better defaults), like speed limit, and separate values for acceleration vs braking effectiveness. The wheels could also use like 50% more torque.

photo
1

Upvoted the thread, I hope keen gives some response soon

photo
photo
4

I think the problem is how small the planets are. If they were scaled up by a decent magnitude, there would be more wheel-friendly terrain. I do not see a single reason to build a rover so far (I'm playing the QuickStart Survival). My first mission was on top of a mountain... why would I bother with wheels when I already have atmospheric thrusters available?

photo
2

Thank you for your feedback. It has been collected and passed on to the developers for discussion.

Cheers,

Arron CM

photo
1

On Verdure, I tried to take the cargo truck and abandoned it after a few hundred meters, because I couldn't get to go where I wanted to go, i.e. the free scrap site. I also thought the braking was unrealistic. That thing weighs over 20 tons and the brake flips it onto its nose! There may be ways to control a wheeled vehicle better but I have yet to find any of them

As for the planetary terrain, artists tend to opt for dramatic features which leads to selecting exaggerated verticals and terrain that seriously unfriendly towards wheeled vehicles. SE1's Lost Colony world also displays this problem. If the terrain cannot be fixed, the wheels need to be both stronger and stickier.

The I went to Kemik and decided to take on a Bulk minerals contract from Dustspire Junction, so I built a copy. The thing, empty, can't even go uphill on the rolling terrain. Admittedly, my expectations were set by driving buggies on Mars in SE1.

So, two suggestions. First, implement regenerative braking. Hitting P while moving would not lock the wheels but instead convert forward momentum into a charging current at a rate proportional to the acceleration setting for the wheels, perhaps 30%, or perhaps allow the scroll wheel to make it more or less aggressive.

Second, implement something to allow the the wheels to stick to the ground better, allowing vehicles to climb steeper slopes.

photo
2

In my (more recent) experience, the biggest problem in some terrains are the numerous boulders and how close the trees are standing. The acceleration of the wheels can already be increased for better hill climbing, but then the problem of high wheel power settings in SE1 will return: An unloaded and therefore light rover will catapult itself into a nearby obstacle with a short tap of the accelerator pedal.

My suggestion for redefining the acceleration parameter: The rover tries to maintain the acceleration in any terrain, using as much power as necessary. Meaning if it accelerates by 1 m/s² on level ground, it will also accelerate by 1 m/s² uphill. To achieve that, it needs to take more power from the battery. There are two bottlenecks that will limit this behavior. The first is the available power from the battery or whatever power source is used. The second is what the motors can handle.

About the "stickiness" of the wheels, also called friction by engineers, I suggest using some high but realistic-ish limit. I did some research and found that high end sports cars typically need a distance of 90 feet to stop from 60 mph. That amounts to a deceleration close to 14 m/s² or 1.4 g in 1g gravity, or a friction coefficient of 1.4

This was measured in cars that have presumably top of the line braking systems, and the tests were done on asphalt which offers good grip. I suggest limiting the friction coefficient of wheels in SE to 1.5 under optimal conditions. Given enough torque of the wheels, that would be enough to climb a slope with a 55 degree angle.

photo
2

Update: I have tried using a rover again today. Braking works reasonably now, I'm getting something like 9m/s² deceleration when braking in the sand on Verdure. The rover also has a tendency to brake more effectively on the wheels that are loaded lighter, when loaded asymmetrically. This will need a bit more study.

Also, I agree with others in this topic that the terrain is a bit too rough. We need more areas with smooth terrain. And that is coming from someone who likes the Alien planet in SE1.

photo
photo
2

Yeah, Verdure as it currently is don't provide much passable terrain to drive on. It's a nightmare to get anywhere; too many impassable mountain, deep cayon, random pits and dense forests almost everywhere. Flying is the way to go as it is.

Edit: And we don't even have water yet; the most driveable parts of the planet are future lakes and seabeds.

photo
1

All of them are like this though, the Kemik moon is impassable, I couldn't drive my Grover D8 ten feet without being completely stopped by some spiky voxel rock formation. I've tested them on each planet and moon now and in general there is little enjoyment in driving anywhere as is.

Step one is the physics engine needs to start treating these rovers as though they have actual mass, half the problem is that the suspensions don't travel at all and the counter force is so great that half the time only one or two of the wheels (out of 6 for the Grover for example, which is a 245 metric ton rover) is actually touching the ground. This makes the full propulsion only come from those wheels in contact with the surface which makes the overall propulsion of the rover extremely underpowered. The suspensions need to travel such that the wheels can stay in contact with the ground more, and the vehicles need to react like they're as heavy as they are and not bounce intot he air every time they hit a pothole.

It should simply be that you hit uneven ground the buggy should remain fairly level almost like it's on a road with the suspension countering the uneven surface with maybe some temporary tilt if it's larger. If it a tree, the momentum of large vehicles should knock those things out of the way, barely slowing (245 tons against a tree = snap and maybe a scratch and dent depending on the speed) The interactions of ships/rovers vs trees should depend heavily on the mass of the vehicle and the size of the tree. I've been bounced/slowed by saplings.

photo
photo
2

Agreed, not enough planning went into verdures landscape, when it comes to vehicle accessibility. They needed to at least designate a network of "trails", even if there wasn't a dirt path across regions of it. That way Players would at least have some direction to navigate/explore towards from most locations; including the starting area.

photo
2

It is supposed to be a new star system to discover. Existing "trails" might not be very believable. What I suggest are overall flatter landscapes with plains where you can go considerably faster without crashing your rover. Closer to SE1. It seems Keen has put more thought into a pretty landscape than a settler-friendly one.

photo
2

"Settler friendly" what does that mean ? By definition a settler transforms his environment to suit his need, especially engineers. Game is about exploring and fighting in a distant stellar system, not hiking in a clean parc. It would be a complete waste to downgrade the planet generation instead of giving the players adapted tools to alter environments, which would strongly enrich the gameplay as well.

photo
1

"trails" don't have to be visible, only functional. aka sections of terrain that is mostly flat-ish, stringed togethor.

photo
photo
1

It would open the possibility for missions to create a trail from a station to a nearby outpost, assuming outposts ever become persistent feature in the landscape.

photo
1

currently driving from the classic survival spawn to Valas Reach (I think I spelled that wrong), and while it is super frustrating, it really isn't that different from trying to drive through the mountains. The terrain itself does need some work, and rocks need to not throw your vehicles into the air, but it really isn't all that bad and they're certainly still useful.


I do think that dropping the center of the wheel, ie. portal axles, would be a great idea though.


On the road front, I made a little two drill miner with no storage and used that to make a 2.5km road w/ 800 meters of elevation change down one of the cliffs into a valley ... and while it was a PITA, it did work and was effin beautiful.

Leave a Comment
 
Attach a file