Ship design/construction/setup ideas for most efficient science collection?

[tfabris]

I'm wondering if there are any interesting ideas out there for the most efficient way to collect Science, in terms of the way you build your ships? Here are some things that I have done, I'm wondering if (a) there's anything I haven't thought of, and (, if anything I've done here is novel:

- All science equipment, including a transmission antenna, are mounted on a rover, allowing me to land my lander at the intersection of a few different biomes, pop the rover off, and then run around and collect data on all of the nearby biomes. Solar panels and batteries galore, to allow for lots and lots of transmitting, since I just leave the rover on the planet when I'm done. I can collect all the science data from space on the way to my destination, while the rover is still attached to my lander. I have to make sure to get all the space science on the outbound leg, because the rover's not coming home with me and thus neither are all the experiments attached to it.

- Rover design is longitudinal, with all the fragile experiment parts along the centerline, and the wheels extended out on long booms, so that even if the rover tumbles in a low-g environment, the experiment bays are not damaged. (Of course the rover is designed to run fine even when flipped upside down.)

- Bonus: With a transmitter antenna on the rover, I can do EVA Reports and surface samples without having to leave my seat. So I can do all science experiments, an EVA report, and a surface sample, all at once, and transmit them all at once.

- External command seat on my main command module/lander section (in addition to the command capsule that's already there), so that I can EVA, move to the external command seat, and be able to do EVA reports *and* Crew reports on my landing approaches or takeoffs without leaving the seat. I haven't actually checked to see if I'd survive re-entry if I hit atmo at reentry speed, so I move back to the command capsule in those situations. But for the Mun or for situations where you're flying around in low atmosphere, it's great.

- Assigning an Action Group Key that contains every science experiment observation, plus a Crew Report. So I just press a single key and the results of all the science experiments pop up on the screen, and then I hit TransmitTransmitTransmitTransmitTransmit. Also add an extra key for just the Materials Bay because I have to transmit that one so many more times than the others. Eventually, as the reports start approaching zero value, I'm hitting TrasmitDiscardTransmitDiscardTransmitTransmit, transmitting only the valuable experiments, to save time and battery power.

Things I haven't figured out how to do:

- How to add Surface Sample and EVA report to an Action Group key. I still have to use the mouse to do those, even though I can do all the other bits of Science with one key.

- How to use keyboard commands to transmit samples, or perhaps a way to automate "transmit all samples all at once, skip all the dialog boxes".

- How to keep the transmitting antenna extended all the time. I can assign an action to extend it, but then when it's done transmitting some data, it retracts again anyway. I'm trying to avoid that moment when it's folding back up and I press the Transmit button on my experiment, and nothing happens because the antenna was folding up.

- How to figure out when you'e driven your rover past the boundary into another biome, without actually running a science experiment.

- How to just Make The Lander The Rover, by adding wheels to my lander, so that I don't have to do the whole "separate the rover" thing when I get to my destination. All of the wheels are too fragile for my big complicated landers which are loaded with fuel and batteries and solar panels and experiments. The wheels just break when the lander sits upon them or moves much at all. I can repair the wheels, but it's useless if they keep breaking every ten meters. I can get the "giant" wheels to work without breaking, but they have a tendency to rip off of my lander body, making repair impossible.

Any other interesting ideas?

[codepants]

How to just Make The Lander The Rover, by adding wheels to my lander, so that I don't have to do the whole "separate the rover" thing when I get to my destination. All of the wheels are too fragile for my big complicated landers which are loaded with fuel and batteries and solar panels and experiments. The wheels just break when the lander sits upon them or moves much at all. I can repair the wheels, but it's useless if they keep breaking every ten meters. I can get the "giant" wheels to work without breaking, but they have a tendency to rip off of my lander body, making repair impossible.

Sounds like your lander/rover (call this an LR from now on) combo is too big. For planets with atmospheres I would have the LR detach from a transfer vehicle while in orbit, then de-orbit itself and use parachutes to land. If you need more than an Oscar-B or two, it's too heavy. The science experiments are comparatively lightweight and you should be able to make an LR with pretty much nothing but them and some wheels. Without a lot of batteries you can use time acceleration to increase your energy intake from solar panels. It's not very convenient, but it's more convenient than having billions of stages.

For planets/moons without atmospheres I'd again think about how big you need everything to be. I'd shoot for having just enough delta-V to de-orbit and land.

Once you've landed you could just de-couple the engines and fuel tanks to make things easier on your wheels.

[Geschosskopf]

Although others disagree, I strongly believe that transmitting is a bad idea for everything except crew reports (which have 100% efficiency). You can't avoid transmitting EVA reports during a long trip but you should never transmit surface samples, Goo, or Jr. Bays.

The most points you will ever get for an experiment is returning the 1st 2 times you run it. This accounts for like 85% of the total available points for that experiment at that biome or orbital location, and it's a complete waste of time to go back for the remainder. So, have 2 copies of each experiment for each zone you plan to fly through and return it all. This makes for missions that are limited by the number of science parts you can get off the ground and land safely. But you can go, grab the points, and come back without having to think too much.

The alternative is to have only 1 copy of each experiment and never return anything. Instead, you run/transmit each experiment in each zone over and over until you get 0.0 points for it, then move to a new zone. The reason you should never return anything with this strategy is because this relies on getting all available points. If you ever stop transmitting to return a sample, that stops you from getting more with that experiment on this flight, so you have to go back for the rest. This makes for missions that can visit as many zones as you have fuel and electricity for, with a lighter ship than the above. But there's a lot of right-clicking involved.

Either way, there's no need for getting complicated and over-engineering the ship. Just do what you have the skill and parts for, then buy some more parts and do something else.

As to your questions:

1. You can't make action groups for actions taken by Kerbals because Kerbals don't appear in the VAB. Maybe that mod that lets you create action groups while in flight would do that, but I've never used it so can't say.

2. Right-click on the antenna and select "transmit data". This will transmit everything you have on board at the time.

3. You can't interrupt the animations of the antennas. Sorry.

4. Look at the map.

5. If you put both legs and wheels on the lander you can do this. Land on the legs, then retract them so it sits on the wheels. Or you can use Infernal Robotics to hinge or rotate the wheels into position. Or, even better, you can not build such monstrous things when there's really no need for them.

[tfabris]

The alternative is to have only 1 copy of each experiment and never return anything. Instead, you run/transmit each experiment in each zone over and over until you get 0.0 points for it, then move to a new zone.

Exactly, that's what I'm doing. You say that's a bad idea, but, assuming that you hit more than one biome with that experiment-laden rover, doesn't that mean that you can get more science points per mission that way?

The reason you should never return anything with this strategy is because this relies on getting all available points.

That's what I thought I was doing...

If you ever stop transmitting to return a sample, that stops you from getting more with that experiment on this flight, so you have to go back for the rest.

I'm not sure I understand that. Can you explain that in more detail? I thought that pressing TransmitTransmitTransmit over an over until it reads 0.0 points (which is what I'm doing) was the same as returning the sample once.

Right-click on the antenna and select "transmit data". This will transmit everything you have on board at the time.

Nice! Doesn't save me any mouse clicks though, does it? I mean, before I do that, I'll still have to press KeepKeepKeepKeepKeep on all the dialog boxes anyway, right?

Look at the map.

The maps on the KSP wiki are at a low resolution, and aren't lined up with a visible-color map reference. So I can't clearly see (from my rover's position on the ground) exactly how far I have to drive to cross a boundary into another biome. I try to land as close to the boundary as I can, so that I don't have to drive very far, but I still can't get it close enough.

Here's an example: I just did a Mun mission where I landed near the boundary between the East Farside Crater, and the east-west canyon that runs eastward from that crater's eastern edge. Going by the map in the KSP wiki, I thought I had absolutely nailed the landing, I thought I was in the area that would be defined as "crater" and would have to drive only slightly eastward to hit the canyon. Instead, I landed, and I was already in the canyon biome, and had to drive about four kliks westward to hit the Crater biome. And also, along the way, managed to find a tiny patch of Midlands that wasn't even on the map. So although "look at the map" gets me in the neighborhood, that neighborhood is still many kilometers wide and involves a lot of driving. The only way, on that trip, that I could tell that I'd crossed biomes, was to run an experiment. The terrain didn't look any different.

If you put both legs and wheels on the lander you can do this. Land on the legs, then retract them so it sits on the wheels. Or you can use Infernal Robotics to hinge or rotate the wheels into position. Or, even better, you can not build such monstrous things when there's really no need for them.

Even when the wheels are intact, I can't get very far on them before they break again, so the legs thing isn't helping me.

And yes, my lander is large and monstrous. But it's also glorious. After a mission, I can fly back to Kerbin and land vertically, on my landing legs, in front of the space center, 50's-sci-fi movie style. That makes it worth it. :-)

[Geschosskopf]

OK, I'll try again....

There are only 2 strategies for doing science: Recover-Only and Transmit-Only. Neither one requires much of a ship to do. Transmit-Only yields a handful more points per mission but requires extensive micromanagement of boring, repetitious run/transmit/run/transmit cycles. Also, it's only really possible once you get solar panels. Recover-Only has none of these drawbacks in exchange for not getting quite as many points.

The thing to avoid is mixing these approaches. If you start out doing Transmit-Only but change your mind and decided to return with a sample after transmitting the same thing a few times before, you lose points and/or have to make a return trip. This is because you can't transmit any more that mission once you decided to keep a sample.

I submit that if the tedious repetition of the Transmit-Only strategy is bothering you so much you're counting mouse clicks, then perhaps you should really be pursuing a Recover-Only strategy. You only use each science part 1 time during the mission, never transmit anything, and come home once all experiments are full. You can spend like 99% of the mission either doing maneuvers or warping, and only have to mouse click once for each science part on the ship. Thus, each mission takes a minimum of your own realtime, so you have rapid turnaround and get more missions in, and thus more science, in each gaming session than you can with the Transmit-Only strategy.

As to the different biomes on Mun, the ground usually changes color and elevation both as you cross the border. Of course, the border areas fade into each other so as to look like real terrain, so you have to go until you're sure you're on a significantly different color of dirt and/or at a significantly different altitude from where you started.

But I have to ask, why bother? Going to Mun just once, twice at the most, should enable you to reach Eve and Duna assuming your couldn't have already. Do a flyby of the big planets and land on their moons. Once you do that, you'll never go back to Mun again because it's chump change compared to what you can get at other planets.

When you have wheels that start out OK but break during use, it's because either 1) you were going too fast or 2) you don't have enough wheels for the weight of the vehicle.

[tfabris]

There are only 2 strategies for doing science: Recover-Only and Transmit-Only. (...) The thing to avoid is mixing these approaches.

Aha, gotcha. Now I see what you're saying, and I agree, thanks!

Transmit-Only yields a handful more points per mission

If transmitting everything allows me to use a rover to cover multiple biomes per mission with only one set of experiments (there are some areas where you can nail three or four biomes within a few klicks of each other), I would argue that transmit-only can theoretically yield quite a few more points per mission, right? Or am I misunderstanding how that works?

Also, it's only really possible once you get solar panels.

Agreed completely. I aimed straight for solar panels and batteries in the tech tree, doing Recover missions up until then.

I submit that if the tedious repetition of the Transmit-Only strategy is bothering you so much you're counting mouse clicks, then perhaps you should really be pursuing a Recover-Only strategy. You only use each science part 1 time during the mission, never transmit anything, and come home once all experiments are full. You can spend like 99% of the mission either doing maneuvers or warping, and only have to mouse click once for each science part on the ship. Thus, each mission takes a minimum of your own realtime, so you have rapid turnaround and get more missions in, and thus more science, in each gaming session than you can with the Transmit-Only strategy.

I see your point there. Additionally, the keep-only strategy alleviates the need for a rover at all, and there's a lot of complexity involved in the rover, batteries, solar panels, antenna, etc, required for the transmit-only strategy. So yeah, I totally see your point.

But I have to ask, why bother? Going to Mun just once, twice at the most, should enable you to reach Eve and Duna assuming your couldn't have already. Do a flyby of the big planets and land on their moons. Once you do that, you'll never go back to Mun again because it's chump change compared to what you can get at other planets.

I think perhaps, I didn't realize at first that the Mun and Kerbin biomes yielded fewer science points than the other planets. Somehow I had it in my head that scouring the nearby biomes would yield more points simply because there were more biomes to explore. Also, I was deliberately trying to climb the tech tree while staying in my local neighborhood and using the biomes as an excuse to explore Kerbin and the Mun in detail. I was also having fun refining my data collection technique and doing interesting updates and revisions to my lander and rover to make them incrementally better. I have a lot of fun in the design screen, fiddling with every little detail. I had my sights set on going to Duna later, after I'd really refined the rover and lander designs.

But you're right, if it was just points that I was after, I should have tried for the more distant planets sooner.

[Vector]

There are only 2 strategies for doing science: Recover-Only and Transmit-Only. Neither one requires much of a ship to do. Transmit-Only yields a handful more points per mission but requires extensive micromanagement of boring, repetitious run/transmit/run/transmit cycles. Also, it's only really possible once you get solar panels. Recover-Only has none of these drawbacks in exchange for not getting quite as many points.

100% agree and to me this is one of the weaknesses of the science system that I hope gets fixed in a future version.

To address the OP question, you want a mission that is "efficient" ... in terms of what? Ship mass? Labor? Number of launches? A rover may be efficient in terms of number of vehicles, but it's going to take a hell of a long time to drive everywhere so it would be inefficient for your time spent. Launching a carrier that separates into a fleet of probes may be efficient in ship mass or number of launches, but not efficient in mouse clicks for all the transmitting...

[KerbMav]

I think - it is already announced and I believe they will tweak the numbers accordingly - that return mission will be more profitable in future versions and transmitting just some kind of down payment to make bigger missions possible.

As of now, as few experiments and sensors as possible, a few antennas to reduce time spent transmitting and big batteries as buffer for solar panels are the way to go - IF you are aiming to unlock the tech tree as fast as possible.

Because players like big numbers, maybe transmitted experiments should be added into a flight log to be read after recovery: "transmitted science points during whole mission: x".