My take on science data/transmission

[Atoning Unifex]

I've seen a bunch of suggestions on science and thought I'd add my own opinions. My thinking was specifically centred around how (in 0.22) I can't dock a lander with a station/mothership and transfer the results over for physical return. I also dislike the data transmission loss. I don't feel it's realistic. I don't doubt some/all of these have been suggested before but maybe not quite in this way!?

1) Split collectible info into physical and data components, so that you can only transmit the data and have to return the physical.

2) Treat both types as a resource, like electric charge, fuel etc, that is shown in the top right. Craft would need sufficient free physical and data storage to hold collected science (options are here for new parts). Transfer between docked craft could be handled like fuel transfers. Physical resource could even have weight to simulate returning rocks. I really like the idea of having to manage data and sample storage and plan ahead, and having storage and computing components to add to your craft.

3) Data transmission should be lossless but there should be maximum ranges for each type of transmitter, and considerations like more power for larger dishes (which i know is already in) and line of sight (you can't transmit through a blocking planet!!).

4) Allow transmission of data between craft. I envisage a lander with a low power transmitter sending data up to a craft in orbit that has the power to then return it to Kerbin. This does present a problem of having two distant ships active at once though. Perhaps a system that pairs transmitters together and can temporarily have both craft active during transmission only. I dunno if this is easy/possible but it feels a more intuitive way of performing data management.

There you have it! Opinions?

[Hilvon]

1) - really good idea.

2) Good ideas too. Some basic capacity for storage may be added to capsule itself (like it stores reports and samples now) and it may very well be limited and expandable by other modules (I also mentioned it in my post you got ahead of)

3,4) Data transmittions may have some data loss due to connection errors, static, interference. But that should be data loss, not sientific-value loss. Moving data from one craft to another would be good. Even if you have to dock them together for transfer (like you have to do with fuel transfers now) That makes even more sense if you want to transfer phisical samples too.

And Mapping connection through other ships may be managed. In fact RemoteTech mod implements it right now.

[Atoning Unifex]

1 is a great idea yeah but I admit to have borrowing it from another suggestion I added it to this post so that the rest made a bit more sense!

2 was my main point to make and you got my exact drift - yeah some storage in command pods with options for expansion.

3 & 4 I personally think data loss should not be an option - yes there may well be interference in data streams but you can simply send the data as many times as you need until it all gets received correctly. I guess I'm not too bothered as long as I don't lose scientific data! I have seen but never used remote tech, I did wonder if it did something similar. Maybe I'll check it out...

[John Nowak]

3 & 4 I personally think data loss should not be an option - yes there may well be interference in data streams but you can simply send the data as many times as you need until it all gets received correctly.

I'm inclined to agree -- this sort of technology was available to consumer goods around twenty years ago.

I'll vote for 1) as well. The whole question of transmitting vs. return needs to be rethought, I think.

[Skyrunner27]

I always thought of transmitting physical data like taking photos of the object and transmitting that.

[UberFuber]

1 is a great idea yeah but I admit to have borrowing it from another suggestion I added it to this post so that the rest made a bit more sense!

2 was my main point to make and you got my exact drift - yeah some storage in command pods with options for expansion.

3 & 4 I personally think data loss should not be an option - yes there may well be interference in data streams but you can simply send the data as many times as you need until it all gets received correctly. I guess I'm not too bothered as long as I don't lose scientific data! I have seen but never used remote tech, I did wonder if it did something similar. Maybe I'll check it out...

Remote Tech data transmission simply asks "Do I have a data path back to Kerbin?". If you have, you can transmit as much as you want without worrying about additional loses (transmission lost in current science system still applies).

I always thought of transmitting physical data like taking photos of the object and transmitting that.

Not necessarily. Some other data that can also be "transmitted".

1. Temperature, Pressure, Gravity, and Magnetic sensor data - Obviously, just numbers.

2. Chemical composition - Assuming on-board mass-spectrometer to determine the elemental composition.

In the end, ALL scientific informations are data, the only difference is that what kind of equipment you need. The way I see it, each science experiment can either be local measurement (like atmospheric temperature, pressure, local gravitational and magnetic anomalies), or samples (chemical compositions). Local measurement requires that you get the scientific instrument there to do the measurement, samples requires that you either bring said sample back to Kerbin (with assumption that Kerbin has all the scientific instrument you have unlocked to perform science on said sample) or you bring the labs (potentially multiple heavy equipments) to where the sample is (or even somewhere in between, park the lab on orbit, and just send down samplers to get sample up into orbit).

[Atoning Unifex]

Not necessarily. Some other data that can also be "transmitted".

1. Temperature, Pressure, Gravity, and Magnetic sensor data - Obviously, just numbers.

2. Chemical composition - Assuming on-board mass-spectrometer to determine the elemental composition.

In the end, ALL scientific informations are data, the only difference is that what kind of equipment you need. The way I see it, each science experiment can either be local measurement (like atmospheric temperature, pressure, local gravitational and magnetic anomalies), or samples (chemical compositions). Local measurement requires that you get the scientific instrument there to do the measurement, samples requires that you either bring said sample back to Kerbin (with assumption that Kerbin has all the scientific instrument you have unlocked to perform science on said sample) or you bring the labs (potentially multiple heavy equipments) to where the sample is (or even somewhere in between, park the lab on orbit, and just send down samplers to get sample up into orbit).

This is an excellent point. So we should be looking at a balance of being able to get the appropriate analysis tools (assuming they are implemented!) to destination compared with bringing physical samples back to kerbin. Then instead of having a %loss due to transmission we have a % effectiveness of certain tools to extract data from physical sources. So perhaps you could get 90+% data from a lifeless planet like minmus, purely based on local analysis, but with a more complex planet with atmosphere, like laythe, you would have to bring back samples to allow proper intesive searches for life and other substances outide the range of your tools...

Maybe that's overcomplicating things....?

[Eric S]

Maybe that's overcomplicating things....?

Sadly, as much as I would love this kind of science system, I have to think this. Something like this would provide benefit for manned missions, as some information could be gained directly through observation, and just about all types of information would require less equipment if the equipment only has to take the measurements and doesn't have to deal with acquiring the sample and moving the sample from instrument to instrument. One way trips would need heavier equipment because they'd need to do the analysis in place, etc. It would also allow you to separate out consumables from the actual equipment, so if you plan running a materials science experiment once, you wouldn't need as much consumables as you would if you were going to run it 10 times, but you also wouldn't need to duplicate all the equipment 10 times.

In the end, however, I realize that if they were to design the game exactly how I would design it, I'd probably be the only person playing it. Playable game mechanics tend to trump realism, and I think they should. The current retransmit-until-no-more-science game mechanic is broken, not fun, etc., and it's getting fixed. Without knowing exactly how they fixed it, all this discussion about how to improve the science system is purely theoretical, as we don't know how much of it will apply to the 0.23 science system, where that game mechanic is getting fixed. This isn't to say that 0.23 will be perfect and need no improvement.

[Atoning Unifex]

Yes, you are totally correct. It's easy to slip into the the-game-should-do-this mentality, which I admit I have done. Pragmatically I know that, no matter what I want from this game, I will be thoroughly enjoying it's final derivation. I just wished it would do it 'this' way....

[Hilvon]

3 & 4 I personally think data loss should not be an option - yes there may well be interference in data streams but you can simply send the data as many times as you need until it all gets received correctly. I guess I'm not too bothered as long as I don't lose scientific data! I have seen but never used remote tech, I did wonder if it did something similar. Maybe I'll check it out...

Well... Here my opinion may really biased, because I have a bit of secon hand (literally "second" as in personal conversation with person with first hand) knowlage of how communications work.

The problem with data loss is not just the interferense and other data damage, but is "Here on earth you ususally have no clue what data is correct and what got messed up".

Sure, there are things like "triple copy", Hemming coding, Checksumms. But they at best let you identify damaged packages. And it's hard to request replacement for those, because the source may go offline, or out of comms' range, or have the ata flushed already to store some other data. And you basically have the same difficultyes with getting your correct request up to the sattelite, with an exception - you have no human brain to work with those errors on the far side.

[rkman]

I agree about ship-to-ship transfer of science data and samples, but although i would like ksp science to generally be more realistic, making data transmissions lossless would remove challenge from the game (no/reduced need for return missions). Something like Mun rocks should definitely yield more science when returned to Kerbin.

Here on earth you ususally have no clue what data is correct and what got messed up".

...But they at best let you identify damaged packages.

If that were the case most of the science that NASA and ESA do would be useless.

At best they can do error correction, and that usually works just fine.

The "data loss" in ksp as it is has nothing to do with radio transmission physics/technology, it is just an incentive to return the ship to Kerbin to get more science.

I can imagine it is very tricky to implement science in a game in such a way that it both realistic and adding something to the gameplay.

Edited December 11, 2013 by rkman

[peadar1987]

I don't really like the "data loss" mechanic, I'd prefer if you got a fixed amount of science for transmitting, say the temperature or gravity readings, another for in-situ analysis of stuff using the science bay or lab parts, and yet another for actually returning samples to Kerbin. If this was coupled with some parts, like a mobile science lab, needing Kerbals to operate it, this would be awesome

[Atoning Unifex]

The "data loss" in ksp as it is has nothing to do with radio transmission physics/technology, it is just an incentive to return the ship to Kerbin to get more science.

Ok so on this basis how about we find alternative incentives to return to planets to gather science? The obvious one IMO is location based science; Have many regions to planets that give different science rewards - I think the Mun already has this but there should be many more regions. Another suggestion, from another post I saw, was that there should be continuous science, so you measure temperature over time, for example.