Too many regolith missions at Alpha Base in r4544, Mission Planning #130
Comments
@shirishag75 , thanks for your observations. Chances are 4 of them are doing a Did you notice any on-going Another case is that 4 of them are on a Mission but due to running out of resources earlier than expected, they are being routed to go to another settlement that is closer. I'll double check and run some observation tests on them. In future, we'll enable the use of |
You can type in Collecting regolith gives them raw resources and is vital to everything they do. That's why there's a drive to go get more regolith. But it's also based on supply and demand. We could allow player to manually adjust the supply and demand curve on a particular resource, e.g. ice, regolith, etc. |
r4546 2018-11-13 Note: this builds addresses issue #130 ## NEW 1. Add getStartingTravelledDistance() in VehicleMission 2. Add vehicle name and type, mission phase, total est. distance and distance travelled to the inquiry on 'mission' in ChatUtils. ## CHANGES 1. In people roster, add # of people on a mission in ChatUtils. 2. Rename "Outdoor" to "EVA Operations" in ChatUtils.
|
Maybe it would make sense to separate mission inception and coarse mission planning from the mission. I thought about it a few times and everytime I see a mission getting not approved again and again I think its a waste of CPU cycles like it is now. It makes more sense when a mission only gets created if it is in fine planning stage and approval about 80% sure. |
|
That does and would make sense as it would reduce the statements to the log which would reduce confusion and make the logs much more readable thank you. In addition to above, we could also some statements like in different bases (sharing an example below) - On Alpha Base These will be sort of reports which would also tell what has been happening. Maybe in future we could also have it period-wise to compare and all but that would be in future. |
mokun
changed the title
r4670 2018-12-14 ## ISSUE 1. #20 2. #130 ## Wiki 1.https://github.com/mars-sim/mars-sim/wiki/Missions ## CHANGE 1. Modify reviewingPhase() in ReviewMissionPlan to add to the mission score of a particular type of mission based on its settlement objective. 2. Change the keyword for seeing the missions in a settlement to "mission plan" and "mission now". 3. Ask players what level (1 to 3) the settlement objective should be - Add adjusting the weight factor in terms of level (1-3) 4. Add log and increase repair task time in RepairEVAMalfunction, RepairMalfunction, RepairMalfunction, RepairEVAMalfunction.
r4698 2018-12-23 ## ISSUE 1. #130 ## NEW 1. Set mission windows to within 50 sols after the start of a new work shift in TaskScehdule. - Add MISSION_WINDOW - Add checking if the time is during the mission window in takeAction() in Mind. - Add isAtStartOfWorkShift() - Add isAtStartOfAShift() for ON_CALL work shift.
|
I've since adjusted the probability and we have much fewer Regolith Missions now in order to give other type of missions the chance to start. DigRegolith is a task that every settler can undertake right outside their settlement. This makes me question why we need Regolith Mission in the first place. Currently, Regolith Mission is not as helpful as Mining Mission. Mining Mission can retrieve tons of valuable minerals that are useful and can be made into many type of amount resources and parts. However, I believe that Exploration Mission must first be done to scope out the region prior to a Mining Mission. At some point, we'll differentiate what Regolith Mission can bring tot he table and why the settlers need to go some place else to get what type of regolith, etc.. Let me know any questions/suggestions/new ideas that you may have, with regard to the use of regolith here. @Urwumpe , do you have anything in particular to add and comment on mining and resource extraction ? |
|
@mokun Well, I would make Regolith collection purely scientific. And even then, I am not sure if it makes sense to travel for a day to just collect a bit of regolith. I would maybe replace it by a "Site Prospecting" mission to test useful mining or settlement sites. This would then bundle multiple prospecting technologies (for example, drilling for ice). A nice extension there could be allowing subsets of different sciences for a site and later create only missions that either verify existing scientific results or does prospecting not yet done on a site. For example just to write some fluffy science fiction here: An initial prospecting mission does a very broad but shallow investigation, because the time at the site is limited to three sols, collecting a bit of regolith samples, drills to maybe 10 meters depth and does some limited meteorological research for the site (Maybe like collecting climate data and collecting blown dust samples), while another teams looks for meteorites. Then after that mission returned and scientific projects related to the mission finished, the site is considered more interesting and new science projects for that site are created, resulting in more and more focussed missions, using their limited time on the site for very specific goals - for example drilling to 50m depth to investigate a underground ice deposit. Or collect more meteorite fragments suspected to be of a common airburst event. The initially just numbered site gets named, for allowing longer missions, the site gets a fixed depot with supplies, better habitats, supply missions come and go, scientific missions get longer and longer... and eventually the prospected site develops into a full outpost that is permanently manned. |
Yes let's list out what kind of sciences can be done on a field trip/site.
Yes that's one area we can integrate better than what we have now. Currently, an areology/biology study does trigger an areology/biology field trip. In both AreologyStudyFieldMissionMeta and BiologyStudyFieldMissionMeta, we have the followings for computing the probability of starting that mission : There is a Task called If we want more features, we'll need to hard code them in. |
Not sure - maybe we can at least reduce the amount of hard coded logic there, for example by describing certain missions as a container for procedures being executed at a remote location. We would then just need a way to describe more or less hardcoded procedures of different tasks and phases in a way, that evaluating the whole mission in the simulation is possible. Then, what we would need is a MetaMission, that can delegate its own evaluation - the score of a mission would then not just depend on itself, but also on scores calculated by the procedures planned to be executed there. Maybe such a container prospecting mission would even be a good prototype for more complex container missions, since we could realistically assume, that the experiments/procedures done on that site can be freely executed in any order without dependencies. And do the more complicated stuff later, once this works. |
Right. we want to encapsulate the purpose, needs/requirements, duration, properties and products, etc. of a mission in json or input file so that we the developers and anyone wanting to do modding can tweak or build something more sophisticated on top of. |
Yeah we can have settlers making use of any meteorite fragments they collected. |
|
In StudyFieldSamples, it said, Our predecessors did manage to connect the rock sample collection to studying those 3 subjects of science and I can certain add meteorological research to it as well. So the question is, what other specific stuff can do we on a sample return and how to quantify if it's a success in unlocking a scientific mystery or not ? I imagine beaming the lab analysis back to Earth and earn some recognition of their work and may be adding a few achievement points to the researchers that belongs to the settlement that sponsors them ?! |
Yes, exactly. Just use hard-coding for the basic building blocks. This also makes it easier to test everything, since we can limit complexity in the tests a lot. We would still need more complex tests than just unit tests on top, but we can at least ensure that the unit tests work fine. |
|
Now looking at ScienceType class, I just discovered a Collaborative Study involves different subject of sciences : |
What about having "Earth" as location in first place and have potential settlers work there? What would be needed in simulation and UI to handle it? Since we are not interested in what people on Earth do in their free time maybe we could use a different "simulation controller" there and only care what happens during their work hours. Then we would just need about 1/3rd of the CPU resources for simulating the same amount of scientists on Earth as we need for the same number of people on Mars. Sounds like a good trade-off. But that should better be discussed in a new issue then for not getting lost here... what about calling it "Design Idea: A Better Earth"? ;-) |
r5112 2019-10-24 ## DISCUSSION 1. #130 ## CHANGE 1. Rename finishedPhase() to approvingPhase() in ReviewMissionPlan 2. Move sortStringBubble() to MathUtils. 3. Rename MissionPhase APPROVAL to APPROVING. 4. Add "final" reserve word to numerous data members. ## FIX 1. Allow old mission plan to pass - Add checking how long a mission plan has elapsed in reviewing phase. - Revise approvingPhase() even if it's not 100% graded 2. Use the false in getAmountResourceStored() in SettlementTableModel in getResourceStored(). 3. Change from "Mission.phase.approval.description" to "Mission.phase.embarking.description" for
r5147 2019-11-05 ## DISCUSSION 1. #130 ## CHANGE 1. Alter when to change work shift to ON_CALL for mission. Should not change a person's work shift right after the mission has just been approved. - Wait till the vehicle is FULLY loaded and right before it embarks. - This way, a person may enjoy the normal work shift and have plenty of rest and sleep right before the mission - Relocate calling setShiftType() (unless it is not a TravelMission) from requestApprovingPhase() in Mission to performEmbarkFromSettlementPhase() in RoverMission. ## FIX 1. Avoid having multiple vehicle operators - Check if a person or a robot is the vehicle operator in OperateVehicle's constructors. 2. Avoid unneeded reclaiming of lost frames when loading from a saved sim. - Add setting excess = 0 inside the for-loop in ClockThreadTask's run() in MasterClock. 3. Avoid NPE in clearSpecificTask() and addTask() in TaskManager. 4. Avoid cyclic calling of endTask() in endCurrentTask() in TaskManager (also recursively calling of endTask() inside endTask() in Task). 5. Give a person the chance to do things other than the mission's driven task in takeAction() in Mind. - Add mission priority in Mission - Toss the dice to choose between getting a new task or performing the mission, with mission priority as the probability modifier. 6. Avoid null study in getAssociatedSkills() in InviteStudyCollaborator. 7. Insert mission designation in MissionEventUpdater in MainDetailPanel when it's a DESIGNATION_EVENT 8. Avoid NPE due to null ScienceType in - getAssociatedSkills() & getAvailableResearchers() in AssistScientificStudyRsearcher.
r5314 2020-08-08 ## ISSUES 1. #247 2. #265 3. #130 ## PROJECT 1. https://github.com/mars-sim/mars-sim/projects/12 ## NEW 1. Add is90Completed() in Task. ## FIX 1. Correct the time usage in ReviewMissionPlan - Call is90Completed() and add calling return 0 to consume time during the task.
|
It's still an open question as to what factors should be reasonably driving regolith and ice collection missions. We have since done a number of things and experimented with fine tuning how these missions should be triggered. We should close this thread now and ask more specific questions, based on what we've implemented. |
The problem is it seems to be a mess, why are so many people are organizing Regolith Prospecting mission and there is no way to know if they are part of a team or they are going to be doing it alone ?
There is also an issue with number of people being reported but this I guess I should report as
a separate bug -
If 24 are registered and 18 are indoors, where are the missing ones ?
The text was updated successfully, but these errors were encountered: