Some Mining Drone Tests

[splatthecat]

Following on from Sit and Spin Mining Drone? and not wishing to further hijack the thread I have carried out some basic tests on three variations of my small mining drone.

All three are small 5x5x5 ships and consist of the same power, thrust, storage, cargo space and a rail docker. None have any secondary systems slaved to the salvage computer but each one has 25 salvage modules set up in different ways.

• Version 1 - 5 separate beams of 5 modules
• Version 2 - 1 group of 25 modules
• Version 3 - 5 separate salvage computers each with a single beam of 5 modules

Apart from the first test which was cut short due to a scheduled server restart, the three drones where tasked with mining a single asteroid until it was completely gone. They were launched to the same starting position around 50m from the carrier and about 10m from each other. They would then have to fly between 250-400m to the asteroid and start their task. Once the asteroid was completely mined the carrier recall command was given and then the contents of each chest examined.

The tests were carried out in factioned space (x12 mining bonus) on the main multiplayer server I play on.

Spoiler: Version 1

Spoiler: Version 2

Spoiler: Version 3

The total number of blocks gathered per mission are as follows.

Version 1 - 5 separate beams of 5 modules

1. 3631
2. 4249
3. 11472
4. 5266
5. 8389

Version 2 - 1 group of 25 modules

1. 3354
2. 3416
3. 15381
4. 7177
5. 9918

Version 3 - 5 separate salvage computers each with a single beam of 5 modules

1. 3522
2. 4706
3. 13782
4. 7441
5. 8256

Individually, the results make no sense. Variations in asteroid shape and size need to be considered and I dont know if mining bonuses are constant or a maximum (1 block x12 or 1 block x1 to x12?) The average across the 5 missions however is more useful.

Version 1 =6601.4
Version 2 = 7849.2
Version 3 = 7541.4

The single beam version on average performs better than multiple beams. Version 3, while obviously multiple beam is essentially 5 single beams.

In flight, Version 1 tended to reposotion itself after each sweep and I doubt slaving cannons to the salvage array would make this better. Version 3 was less prone to this however it would sometimes sit without firing the beams. Version 2 would more often than not, find a position above the asteroid and carry out methodical right to left sweeps. I would suggest that there was a 65% reduction in movement.

Someone said that multiple beams are currently pointless as only one beam works anyway. Well visually this isn't the case. All beams show they hit and blocks disappear from each one.

Spoiler: All Beams Visually Effective

The methodical sweeps of Version 2 can be seen below. The plateau was created by this and the beam can be seen coming in from above.

Spoiler: Single Beam Sweep

While I admit more testing is required, I am thinking of replacing my Version 1 drones with Version 2 drones. I may also experiment with slaving cannons to the salvage array but this will mean I will have to replace some of the salvage modules with cannon modules, maybe 12 + 12 so I may have to do further tests of this variation compared to the non slaved version.

I hope these tests were of interest.

 

[StarWars1981]

Interesting, very interesting. I'll have to build my own set of mining drones sometime. Should be fun . GJ testing this for all of us, and thanks!

 

[StormWing0]

Your images are broken on Chrome. >_>

 

[Madman198237]

Very good testing methodology. However, you should've simply each one a separate test, so they don't all eat the same asteroid simultaneously. For future reference, I would recommend the same course, but instead of mining 1 asteroid, you send them out for a specific length of time, then calculate how many blocks it can mine in a set amount of time.

Do this for each drone, independently, so you don't have error from drones being forced to maneuver around each other and the possibility of messed-up mining due to multiple drones mining the same block/group of blocks.

 

[splatthecat]

Very good testing methodology. However, you should've simply each one a separate test, so they don't all eat the same asteroid simultaneously. For future reference, I would recommend the same course, but instead of mining 1 asteroid, you send them out for a specific length of time, then calculate how many blocks it can mine in a set amount of time.

Do this for each drone, independently, so you don't have error from drones being forced to maneuver around each other and the possibility of messed-up mining due to multiple drones mining the same block/group of blocks.

That is not a bad idea. I will carry out some more tests tonight an post the results.

 

[Sven_The_Slayer]

Probably the best way to do tests would be to export a sector and mine the same sector and time how long it takes from start to finish. I'm tempted to do this myself as I am really curious to see the results plus comparing things like cannon salvage and maybe pulse salvage as well.

 

[splatthecat]

If anyone wants to carry out their own tests under specific conditions feel free to add your results to this thread or comment on your findings. I'm sure people will find them interesting.

Anyway, I have carried out a further batch of tests. Each drone was sent out individually and tasked to mine the sector. I allowed the drone to fly to its' starting point and started the clock when it began mining. After five minutes I issued the carrier recall command and examined the storage contents. As before, I did five missions each qnd took an average.

Version 1 - 5 separate beams of 5 modules

1. 1954
   
2. 1896
   
3. 1902
   
4. 1933
   
5. 1881

Version 2 - 1 group of 25 modules

1. 2209
   
2. 2283
   
3. 2189
   
4. 2314
   
5. 2055

Version 3 - 5 separate salvage computers each with a single beam of 5 modules

1. 1644
   
2. 1306
   
3. 1398
   
4. 1459
   
5. 1519

This gave averages of...

Version 1 = 1913.2
Version 2 = 2210
Version 3 = 1465.2

The single beam version once again has triumphed so I decided to make a modified version and run the same set of tests as a "just out of interest" comparison.

Version 2B - 1 group of 12 Salvage + 1 group of 12 Cannon

1. 1977
   
2. 1897
   
3. 1801
   
4. 1957
   
5. 1864

This gives an average of 1899.2

It would appear from both sets of results that the single beam drone worked better than the multiple beam versions. Version 2 is more stable in flight and seems to spend more time salvaging than trying to reposition itself. Version 3 was almost as stable but the AI seemed to have to calculate the optimum position for all five beams before firing so sometimes sat doing nothing.

Vesion 2B was an interesting further experiment. With approximately half the number of salvage modules, it was able to gather almost the same number of blocks as Version 1 and was considerably more efficient than version 3.

From these tests, my personal opinion would be to rebuild my mining drones to Version 2. I would assume that these results could be applied to larger drones if that is your preference. Grouped, single beams appear to be better than the multiple beams of an equivalent size. Using multiple salvage computers to make multiple single beams (Version 3) helps with stability but is less efficient in block gathering. Larger drones would probably benefit from secondary effects where space is less of an issue but single beams appear to be better than multiple ones.

Please feel free to expand upon these experiments or share your own findings or analysis of my results.

 

[Madman198237]

Nice. Exactly my thoughts on experimental procedure.

Very good results. Now, maybe we should go about testing more effective combinations? I'll look into doing that, but it probably won't get done today.

 

[Crimson-Artist]

excellent job! might have to rethink my new salvager

 

[Edymnion]

Yup, interesting results. Guess I'll redesign mine.

 

[Edymnion]

Yup, I just redesigned my salvage drones to be a 50 module cube (50 salvagers, 50 cannons), and they flew MUCH better than my old sit and spin happy fun play time rod salvager did.


Random notes from my test:

All drones apparently try to target the same block on the same asteroid.


Still have issues with them deciding to spin like idiots, but it seems mostly limited to when they finish eating an asteroid and don't know how to move on to the next.


So, it seems its going to require a MASSIVE number of salvager drones if single beams are the fastest and most efficient (by the end I was testing out a swarm of 16 drones and it was almost getting to an acceptable, if slow, speed).

And actually, I think I kind of prefer it that way. You can manually pilot a miner and clear sectors out quickly and efficiently, at the cost of flying it yourself, or the drone swarms can (mostly) do it automatically, but take much longer.

Hmmm... I wonder if drones can Mine Sector on overheating ships?

 

[Madman198237]

Not yet, no. They can only mine asteroids for now. But soon...soon (TM)...we will destroy planets and overheated ships in the wreckage of a battle alike.

 

[wanzer]

that be nice on speedy ships that still have they're momentum after they overheated
no more hitting them with stop cannons just send in the drones

 

[Edymnion]

Yeah, but you'd need some pretty hefty drones to really make any use of it, come to think of it.

Most of my overheat scavenging is against pirates, and there's always another wave or three coming in while I'm eating the last one. I don't consider myself to be in the rhythm yet until I'm doing nothing but eating and my turrets are killing everything that gets close. I'm just moving from one overheat to the next, only occasionally firing the main weapons.

Which would mean the drones would have to be out in the middle of a half dozen pirate ships and likely a station at any given time, which means significantly more shielding per drone just to keep them alive than I'd really want to spend on drones.

Now, if we had the option to build a large fighter that could hunt down pirates, kill them, eat them, and gather up loot clouds, that would be pretty awesome. But at that point whats the use of going outside?

 

[Sven_The_Slayer]

Yeah, but you'd need some pretty hefty drones to really make any use of it, come to think of it.

Most of my overheat scavenging is against pirates, and there's always another wave or three coming in while I'm eating the last one. I don't consider myself to be in the rhythm yet until I'm doing nothing but eating and my turrets are killing everything that gets close. I'm just moving from one overheat to the next, only occasionally firing the main weapons.

Which would mean the drones would have to be out in the middle of a half dozen pirate ships and likely a station at any given time, which means significantly more shielding per drone just to keep them alive than I'd really want to spend on drones.

Now, if we had the option to build a large fighter that could hunt down pirates, kill them, eat them, and gather up loot clouds, that would be pretty awesome. But at that point whats the use of going outside?

What I did recently was I jumped in my primary salvager, ordered my fleet to attack a pirate station and as the fleet was busy and pirates were dropping like flies I would eat them up as fast as I could. My salvager is a bit of a tank so I was never at risk of damage.

 

[Edymnion]

What I did recently was I jumped in my primary salvager, ordered my fleet to attack a pirate station and as the fleet was busy and pirates were dropping like flies I would eat them up as fast as I could. My salvager is a bit of a tank so I was never at risk of damage.

Same here, my miner has something like 3 million shields and the main lock on missiles I can't use on Isanths because it vaporizes the entire ship so that there's virtually nothing left to scavenge.

But then thats building a drone fleet for combat while you scavenge, instead of a mining drone fleet while you fight.

 

[Sven_The_Slayer]

Same here, my miner has something like 3 million shields and the main lock on missiles I can't use on Isanths because it vaporizes the entire ship so that there's virtually nothing left to scavenge.

But then thats building a drone fleet for combat while you scavenge, instead of a mining drone fleet while you fight.

I'm not sure how my drones would fair in a combat zone, they are fairly large compared to other drones.

 

[wanzer]

still bothered that I don't get a diamond on all the dropped junk
since the ships move often rly fast tracking the loot that miraculously stay at the same spot while the rest of the ship hurls into space

 

[Edymnion]

still bothered that I don't get a diamond on all the dropped junk
since the ships move often rly fast tracking the loot that miraculously stay at the same spot while the rest of the ship hurls into space

One way to help fix that is to put Stop on your turrets.

 

[Madman198237]

We should take all the research done in this thread (Which is rapidly turning into the Mining Drone version of the Drone R&D thread) and start making suggestions for improvements based on it.

That and continue the research.