Mission Planning

[Excalibur]

Seeing as though Fine Print is currently being integrated in to KSP it gave me an idea on how we could improve the game.

Fine Print has some pretty cool visualisations of target orbits and the like; assuming these make it through the integration in to stock why not put them to use for other purposes such as mission planning? Below is a rough outline on how it could be done:

Create a new scene that is outwardly very similar to the current tracking station. Lets call this the Mission Development Centre (MDC) for the sake of argument. Upon entering the MDC the game is effectively paused and a 'snapshot' of global time, orbits and craft/debris is created. Essentially a simulation, the player can now warp time forwards and backwards (though you would not be able to warp to before the time you entered the MDC). As it's a simulation, it will not have any affect on the passage of time in the actual game.

This is where the Fine Print visualisations come in. Allow the player to create a starting orbit such as a zero-inclination 70x70km Kerbin orbit (or any other around any other body). From here they can add maneuver nodes at will, ultimately producing complicated mission plans such as grand tours at any point in time.

Once a mission plan has been finalised it can be saved for use; when the player subsequently enters the flight/map/tracking scenes these future orbits and maneuver nodes will be displayed and in a similar fashion to some of the Fine Print contracts. The player will have to fly the maneuver nodes accurately enough to reach their next node at the correct time. The crux of this is that unlike traditional nodes these 'mission nodes' will not be affected by changes in your current trajectory and be more like keyholes in space and time to aim for.

I think a system that's close enough in functionality to what's described above will give players a much more powerful tool for mission planning whilst largely removing the need for phase angle calculators; anyone can jump in to the simulation, create nodes and play with time until they get the results they desire. After fleshing this out I had a search on the forums to look for similar ideas and found this - a very similar post from April. Since this thread is dead and my thoughts have added to the subject I think it's fair to highlight simulated mission planning again.

[CaptainKipard]

I'd prefer to have something resembling KER. I plan everything with a spreadsheet and KER. The visualisations would be nice, but I think a simulation is going to far. I don't imagine I'd enjoy doing the same things twice essentially.

Edited December 1, 2014 by Cpt. Kipard

[Excalibur]

I'm a fervent user of KER, would never want to play KSP without it. For me, having hard numbers to work with while trying to work to a given delta-v budget is indispensable.

What I tried to describe in the OP above is a visual system that lets you chain together a series of maneuver nodes, each at a different point in time and space whilst giving you the ability to edit, refine and save them for use in flight. Personally I think that a system that does this visually in-game is the preferred option for most players. As useful as they are, spreadsheets are accessed by dipping out of the game and this breaks immersion.

I don't really see what your point is regarding having to do the same things twice. The simulation is about setting up transfer windows and slingshots, there's no need to fly it or watch a representation do it for you. It would all be done on-rails and without the ability to focus on a vessel, all the player would be doing is setting up maneuver node 'checkpoints' for use as a reference in a later, live mission. You would choose the trajectory you want to follow in the mission, then fly it for real.

The only thing that's potentially repeated is messing around with maneuver nodes, once in the planner and again for real. I think the advantages of such a planner outweigh something as insignificant as having to play with maneuver nodes a bit more - it's something we all spend a lot of time doing anyway.

[Bill Phil]

Huh. Weird. I had this exact thought ( although not with fineprint) a while ago...

Plus, each mission profile can be reused!

And maybe make it a tab at Mission Control.

[Red Iron Crown]

An interactive delta-V map, where you can select the origin and destination nodes and have it compute the delta-V requirements, combined with a delta-V meter and some way of showing transfer windows, would make mission planning much, much easier. I currently use mods (KER and Transfer Window Planner) and a paper delta-V map for this functionality, something similar in the stock game would be very welcome.

I like the implementation of maneuver nodes, but I'm not sure they're the best way to do the sort of high level planning we're discussing. A bit too fiddly, and somewhat repetitive as Cpt. Kipard points out.

[klgraham1013]

I've thought about this as well. Planning and pulling off a complex maneuver like Rosetta would be much simpler with an interactive map to plan the mission with.

[sherkaner]

As wonderful as KER is, I think a simpler tool in stock would be great for less numerically-oriented players who still want to figure some things out before designing and lighting a rocket. I like something like Excalibur's idea, although I agree that strictly doing it with maneuver nodes seems a bit under-powered and fiddly (much in the way that doing it live once you're in orbit is fiddly!). I wonder if somebody clever could come up with a nice intuitive, visual way of letting a player play with orbits and burns in a way where they can see things like the most efficient time to burn to a new orbit, or a good alignment of multiple bodies for a more complex mission.

[Nemrav]

Ah, a planetary orbit sim except maneuver nodes are way more accurate and (since it is its own scene) can handle more of them... Dv for each maneuver and everything...

But what happens if I miss my target , maybe mission planner could perhaps suggest a correction maneuver ?

[sherkaner]

Personally I wouldn't want to see it become a tool to take the thought out of anything. I like the idea of the mission planner just providing fixed reference waypoints. If you miss those though, you're on your own. I'm all for people having MechJeb and KER for more rigorous and "low risk" mission planning, but in stock I rather like the game being very seat-of-the-pants. It would be nice to have a tool to figure out what should be possible ahead of time as you're building and planning.

[Vectura]

I like this idea very much

[John FX]

Personally I wouldn't want to see it become a tool to take the thought out of anything. I like the idea of the mission planner just providing fixed reference waypoints. If you miss those though, you're on your own. I'm all for people having MechJeb and KER for more rigorous and "low risk" mission planning, but in stock I rather like the game being very seat-of-the-pants. It would be nice to have a tool to figure out what should be possible ahead of time as you're building and planning.

If it does become a tool then it is one that you can have the choice of using or not using. It would not fly the nodes or mission for you, just show what you have to do and when, like the current system.

I like this idea very much

So do I. It would allow you to make a flyby mission much more easily for example.

I like the way you do not simulate and fly the mission but instead just build the Dv nodes ahead of time for an idealised mission. Combined with the time shifting to see what is going on when you arrive it could be a good tool.

I assume the OP imagines there will be a `total Dv` shown somewhere?

[Excalibur]

Great feedback!

I did not want to set in stone what the mission planner should be, rather what it could be and open the subject for discussion.

Cpt. Kipard and Red Iron Crown; I understand your concerns of the potentially repetitive nature of the tool described above. The thing is I just cannot think of another way to do it without becoming too complex (and therefore opening up it's use to as many players as possible). To make the proposal more feasible and attractive to the devs I decided to suggest the use of the Fine Print visualizations as they may be implemented in to 0.90 along with the rest of the mod.

Likewise, using a modified version of the current maneuver node system would reduce the amount of work the devs would have to do to get a more powerful mission planning system in place.

To describe how I envision a system like the one I've suggested will work here's an example;

Mission:

Fly-by of Jool with ultimate destination being an Eeloo landing.

Planning:

• Create the starting orbit in LKO (70x70, zero inc.)
  
• Create a maneuver node for the transfer to Jool. The player can warp time and shift the node until they have what they feel is an optimum transfer. This node displays the trajectory to Jool and beyond.
  
• Player discovers that with the planetary alignment they've chosen a gravitational assist to Eeloo is not possible.
  
• Player warps time ahead further until they reach a planetary alignment that results in a successful transfer to Jool and slingshot to Eeloo.
  
• Create a maneuver node in Eeloo's SOI that results in the desired orbit.
  
• Now the mission plan is saved with each maneuver node having it's own location/time stamp.
  

Implementation:

• Before lift off and in the map scene, the player will see the highlighted initial LKO parking orbit and Jool transfer node.
  
• Launch and attain the specificed parking orbit (just like the attain-an-orbit contracts in Fine Print).
  
• Also visible will be the planned trajectory to Jool, the maneuver node at Jool, trajectory to Eeloo and then the final node in Eeloo's SOI.
  
• To follow the planned mission profile the player then has to burn the nodes in sequence and as accurately as possible. Any errors in transfer burns will not change the position of nodes planned in the mission planner. They remain fixed in time and space as windows to pass through. If you burn at the wrong time or with the incorrect delta-v then you will have to make mid-course corrections to hit those windows (admittedly this won't be easy with complex missions). This recreates how real interplanetary missions have to make precise burns to pass through specific points in space and time.
  

There are issues with implementation; using the mission to Jool/Eeloo above as an example:

• How would you 'hold' the initial Jool transfer node in the optimum Kerbin --> Jool departure angle while warping ahead to then find the optimum alignment between Jool and Eeloo for the powered gravity-assist?
  
• Would the targetting system be able to compute the closest approach to Eeloo in the planning screen effectively?
  
• If using maneuver nodes would indeed be too fiddly for this task - what alternatives are there or could be created?
  

Currently there is no way in the stock game that allows you to plan such a mission; mods do help but as far as I know none let you plan for multiple gravity-assists (if I'm wrong I'd love to know) and spreadsheets are very difficult to set up for a lot of players (again if I'm wrong please correct me). If Fine Print is being integrated in to stock in it's entirety and adapting existing assets and code for the purposes described above is feasible, then there may be hope in getting this suggestion taken seriously at some point in the future. I really think that KSP would benefit immensely from a powerful mission planning tool that doesn't require mods or spreadsheets.

If it does become a tool then it is one that you can have the choice of using or not using. It would not fly the nodes or mission for you, just show what you have to do and when, like the current system.

So do I. It would allow you to make a flyby mission much more easily for example.

I like the way you do not simulate and fly the mission but instead just build the Dv nodes ahead of time for an idealised mission. Combined with the time shifting to see what is going on when you arrive it could be a good tool.

I assume the OP imagines there will be a `total Dv` shown somewhere?

That's exactly it. It's an idealised mission plan for you to follow and aim to replicate - the player does all the work in setting up the plan and following it accurately. If you miss the optimum space-time window for a gravity assist by more than a few days there's a real chance you'll miss future rendezvous and lack the delta-v to correct your mistake. I don't see any reason why a total delta-v readout could not be included.

Edited December 2, 2014 by Excalibur
Typos

[Red Iron Crown]

Excalibur, you might be interested in trying Trajectory Optimization Tool, software that allows planning of multiple gravity assists and aerobraking, etc. Not exactly simple or easy to use, though.

I'm definitely on board with the goals of your suggestion, the game desperately needs an easy way to plan interplanetary missions IMO. The use of maneuver nodes has the advantage of leveraging an already developed interface that the player is already familiar with, so I like the idea in that sense. Maybe if the maneuver node interface made being precise a bit easier I would be less resistant to it. Part of the problem is that it sort of requires you to already understand a bit about how to do IP transfers, it doesn't improve the discoverability very much aside from not having to launch a ship to mess around with the nodes.

There are two things needed, IMO:

1. Some way to determine the dV (and thus fuel) requirements for a mission and craft. This means some sort of dV calculator, perhaps simpler than the ones added by mods; and a dV map of some sort. I think a dV map made of clickable nodes for origin and destination is quick, simple and lends itself to easy graphical representation.

2. Some way to convey information about transfer windows. Maybe color gradients on the target's orbit? I.e. when the target enters the bluish area of the orbit line, the transfer dV cost is approaching optimal.

While I understand the desire for more sophisticated tools for gravity assists and aerobraking (and porkchop plots for non optimal transfer calculation), I'm not sure they can be as easily integrated into the core game without it getting into "information overload" territory.