|
Post by Hannes Vilhjalmsson on Mar 3, 2009 17:17:06 GMT -5
This week you will be reading "Activity Planning for The Mars Exploration Rovers" by our own Ari K. Jónsson and colleagues at NASA. This will relate the topic of our upcoming Friday lecture, "planning", with space exploration! Ari will be our guest next week in our Tuesday discussion class, so prepare a couple of good questions to discuss with him.
|
|
|
Post by Helgi Páll Helgason on Mar 6, 2009 4:53:38 GMT -5
My understanding of the paper is that planning is done here on Earth and a list of specific instructions is sent to the rovers for execution. Is this correct?
How are unexpected situations, occurring at execution time, handled on Mars? How reactive are to rovers to their environment? Would the rover continue to drill into a rock if a little green man showed up? :-)
|
|
|
Post by Bjarni Gunnarsson on Mar 6, 2009 18:51:29 GMT -5
How many man hours does it take to program an average day for the rovers?
What if anything would you do differently if you were to send a rover to mars today?
|
|
|
Post by Jón Trausti on Mar 7, 2009 16:17:35 GMT -5
Pretty amazing that this worked out so well.
Is MAPGEN still being used/developed today? And if NASA would send another rover, would it use MAPGEN again?
Were there any large mistakes done that couldn't be fixed because the Rover is on mars? Regarding MAPGEN or the hardware?
One more.. the satellite that brought the rovers to Mars, what kind of system is used to make it head to Mars?
|
|
|
Post by Birna Íris on Mar 8, 2009 9:33:34 GMT -5
How complex was the implementation? For example, how many constraints where "working" in the system at a time? Where all the constraints editable or where some static in the system? And how many actions are available?
Planning vs. Scheduling? What are the main differences?
I am also very curious about the user interface and how it was designed and developed. A lot of emphasis is on the TAP user and as I understand it this person is under a lot of pressure of making the correct decisions at each time. This makes the user interface extremely important.
Another thing I find very interesting is the message sending between Mars and Earth. What kind of equipment is used, how fast do the messages travel, is it in realtime?
|
|
|
Post by Haukur Jónasson on Mar 8, 2009 12:09:43 GMT -5
Quite an interesting article this time. - How would this project have been done differently if the ground-team would have been able to communicate with the MER in (near) real-time, (such as if they were in orbit over Mars, in a Mars ground base, or in possession of futuristic FTL communication technology)?
- What other kinds of projects is this combination of AI automation and human operation useful (and has been used) for?
|
|
|
Post by hordur08 on Mar 8, 2009 14:06:29 GMT -5
1) How advanced is the use of solar energy? Does the lack of energy put constraint on how much research can be done? Do the Rovers function during the night?
2) How much time do you have contact with the rovers at a time and how long and for how long time goes between contacts?
|
|
|
Post by Jon Gisli Egilsson on Mar 8, 2009 14:11:20 GMT -5
The first questions that wanted to ask have already been asked so ...
1) What are the people who worked with you on this project working on now?
2) Were there any flaws in the system which you didn't notice untill the rover was on Mars? If so, can you repair the system from earth with out a big fuzz.
|
|
|
Post by Christian Zehetmayer on Mar 8, 2009 14:27:03 GMT -5
Are there any emergency plans when the rover gets stuck, runs out of battery or other extraordinary situations?
Why is a mixed-initiative system for the rover chosen? Must a human be responsible for decisions?
|
|
|
Post by olafurgi on Mar 8, 2009 18:25:43 GMT -5
Interesting reading.
Has the MAPGEN or a modified implementation fo MAPGEN been used utilized in a project with similar requirements ?
Did difference in measurement units between the MAPGEN and other modules cause any difficulties ? (SI Units vs. Imperial).
|
|
|
Post by Snorri Jónsson on Mar 8, 2009 18:56:30 GMT -5
In several places limited power is mentioned what function of the rover is taking most of the energy?
What parts of the rover are functioning during the night? Just a clock?
|
|
|
Post by steinarhugi on Mar 8, 2009 20:06:55 GMT -5
Since this is anything but a traditional project, I'm interested in the development process. How was it? - How big was the team and how was it constructed? - What methodology did you use? - How did you test it? - How was your architectural design pattern?
In the paper it is mentioned that the mission lifetime was limited. Is that due to high chance of equipment damage or limited funding time?
What is the current "health" status of the rovers?
|
|
|
Post by Hólmar Sigmundsson on Mar 8, 2009 21:11:36 GMT -5
What was your biggest contribution to the project?
In general, how much time did the ground team have to make the plan for the next day?
|
|
|
Post by arnists on Mar 9, 2009 10:12:20 GMT -5
Page 9 line 9 states "indicate a 20-40% increase in science return over the manual approach used in Mars Pathfinder"
By what metric?
What sort of test environment did you run your software against?
|
|
|
Post by Hjalti Kolbeinsson on Mar 9, 2009 15:52:27 GMT -5
What was the expected lifetime of MER and how long was it actually "alife"?
In a system that is calculating so much, isn't it extremely difficult to debug and find if there are some errors in the calculations?
Is NASA currently working on another explorer for Mars or even some other planets?
|
|