Printable Version of Topic

Posted by: Actionman Aug 28 2014, 11:11 AM

Driving at night might be better.

Posted by: djellison Aug 28 2014, 12:42 PM

How would you conduct hazard avoidance in the dark? How much extra energy would you be spending heating the mobility system to allowable temperatures before moving?

And moreover....why might it be better at night anyway? I've attempted to drive across the sand in 40degF heat and 105degF heat....didn't change a thing.

Posted by: Actionman Aug 28 2014, 09:23 PM

You would use infrared time exposers with real time overhead data link. Or just do a short test and compare wheel slippage with the daylight data which you could after the fact..

Posted by: djellison Aug 28 2014, 10:04 PM

The Haz and Navcam's are not IR cameras. They can not see in the dark.

https://www-robotics.jpl.nasa.gov/publications/Mark_Maimone/fulltext.pdf

Their sensitivity is between 600 and 800nm. Thermal IR starts at 3,500nm

What is a 'realtime overhead data link'? Are you talking about returning data from MRO? That only happens for about a 15 minute period, twice a day, and the one-way light time is usually more than the duration of an MRO pass.

Posted by: Actionman Aug 28 2014, 10:54 PM

1035nm MastCam with starlight is good for dune driving.

With just one peak form MRO it would be a short drive

But it could be worth it if the dune dust is more cohesive after it's cooled down.. say some time after midnight.

Posted by: Explorer1 Aug 28 2014, 11:14 PM

If there was any advantage to driving at night at all, it would have been done at least once some time in the last two years, so let's give the planners the benefit of the doubt. They're pretty good at this sort of thing, after all.

Posted by: djellison Aug 28 2014, 11:39 PM

MastCam can not be used for AutoNav.

MastCam will not see anything in 'starlight'

Moreover
-The MRO pass would not be long enough for an in-the-loop driving decision process.
-Earth may well not be visible during a night time MRO pass at Gale.
-The energy required to warm the mobility system to do this would be huge (if not entirely insurmountable)
-What evidence is there that 'cool' sand dunes offer more traction?

Posted by: serpens Aug 29 2014, 02:36 AM

The silt, sand and dust that makes up the ripples seem to be mechanically locked. The surface is absolutely desiccated and empirical measurements suggest that there is no transfer of the miniscule atmospheric H2O to the regolith overnight other than very minor surface frost so there would be no ice cementing if that is what you are positing. Any minor contraction due to temperature drop would in fact reduce the effectiveness of mechanical locking such as it is. The only potential gain is that aluminium would have a marginal increase in tensile and yield strength at the very low overnight temperatures but hardly sufficient to provide mitigation for wheel damage. IMHO it would be pretty senseless to take the risk of night drives when there is no payoff.

Posted by: Actionman Aug 29 2014, 10:59 AM

It will be imperative to cross dune at some point with or with out confidence. I don't see any reason to think a dune walk at night is any more dangers then at day. That should be a one move operation none stop crossing.

Posted by: Gerald Aug 29 2014, 12:00 PM

It's more dangerous, because the ratio of slip versus wheel rotation cannot be monitored in the dark by VisualOdometry.
Hence excessive slip above a previously defined trigger level may not be detected in time.
This increases the risk to get stuck.

Early detection of excessive slip allows tracking back and choosing alternative routes, until a successful crossing is accomplished, if crossing the dune turns out to be necessary at all. South of the entry point there are gaps between the large dunes. On the way to the entry point, there is no need to drive away more than 10m from solid ground or shallow sand.

Posted by: RoverDriver Aug 29 2014, 12:54 PM

Driving at night would be quite expensive due to the amount of heating necessary to keep the actuators warm and cozy. But mostly we could not use VO or autonav. We could just drive blind and use other parts of safety mechanisms to detect embedding but these would be very, very short drives, 10-20 meters at most. And so far I haven't heard anyone suggest that sand on Mars is more cohesive in cold temperatures. If it was, we would have seen differences in driving Spirit and Oppy, besides Curry, in Winter vs Summer.

Paolo

Posted by: djellison Aug 29 2014, 02:45 PM

It is imperative to not do something that is likely to get the rover permanently stuck. Driving into areas with questionable traction without autonav & visual odometry etc is significantly more likely to result in a permanent embedding event that would bring the mission to a premature end.



Just because you don't see them doesn't mean they don't exist. It's been explained why driving in the dark would present an entirely unnecessary risk to the vehicle

Posted by: Actionman Aug 29 2014, 06:16 PM

A night test is not so far fetched.

The electrostatic continuity of the subsoil will be enhanced in a sub -40 degree night environment, compared to the heat of the day... implying the possibility of a increase of cohesiveness in the soil.

Posted by: Phil Stooke Aug 29 2014, 06:33 PM

No energy for actuator heating at night, no light for visodom, slip checks etc. Give it up! The best engineers on Planet Earth are already on the job.

Phil

Posted by: mcaplinger Aug 29 2014, 07:12 PM

I'm not 100% sure this is true, as it was discussed as a backup in case of Navcam failure during development. But it's certainly not assured; Mastcam is not even allowed to be powered on during driving at present.

As has been pointed out, there seems to be some confusion about how night vision works. I know of three flavors: image intensification of ambient light ("starlight scopes"), near-IR with active illumination in the near-IR (seen in some security cameras, baby monitors, etc.) and thermal IR. We can't do any of those with anything on MSL and the longest exposures we could take would be unlikely to even detect the ground. Near-IR works no better than visible for ambient light -- worse, because the sensitivity of Mastcam in the near IR is quite low compared to the visible. The best we could do is use the MAHLI LEDs for illumination, but the arm can't be deployed while driving and in the stowed position the LEDs aren't pointed in a useful direction, and they're not all that bright anyway.

It's an interesting idea that slip properties might change as a function of temperature, but I don't think there's any evidence that this might be true, and I don't think there are any known examples of it for terrestrial analogs.

Posted by: SpaceListener Aug 29 2014, 07:24 PM

Finally, there is no hurry so it is not necessary to risk if the time is plentiful. Besides, as there is no hurry, take the easiest and safest way is the most wise mission plan. According my Earth experience is that the night driving on sand has a little better traction since the sand is little more cemented due to the humidity but in Mars, the humidity is almost inexistent.

Posted by: nprev Aug 29 2014, 08:13 PM

The bottom line based on MSL's capabilities and this discussion seems to be that any benefits (scientific or otherwise) that might be derived from a night drive would be minuscule compared to the level of risk that would be assumed by doing so.

Given that the two words that best describe spacecraft mission managers--esp. those responsible for a Flagship-class US planetary mission--are "risk averse", this just ain't gonna happen.

Posted by: RoverDriver Aug 30 2014, 04:00 AM

It is *very* easy to make suggestions of this level from the comfort of your home, but it is quite different when you are in the hot seat and need to ensure that a $2.5B national asset in a very visible space mission is not damaged, or its mission jeopardized. We have indeed entertained the idea of driving in the dark but mostly to increase the amount of driving we can do in a Sol and the added risk and effectiveness has been determined not worth the trouble.

Paolo

Posted by: serpens Aug 30 2014, 05:41 AM

Would you mind explaining your rationale for enhanced electrostatic continuity with this degree of temperature drop in a desiccated, particulate environment? I regret that my education is sadly lacking in this respect.

Posted by: Actionman Aug 30 2014, 11:43 AM

It wasn't very easy making this suggestion not at all, there seems to be a necessity in the making. You just did a test run in a small dune area (which looked fine to me). So there may not be a scenario where the mission fails or is deemed unassailable. Correct me if I'm wrong but isn't avoiding dune wasting precious time.

Posted by: RoverDriver Aug 30 2014, 01:04 PM

From the outside some of the mission decisions seem puzzling, some even bizarre but there are many constraints we have to deal with. Power, science observations, communication windows (which are set weeks, if not months in advance) which sometimes force us in doing what we are doing. If you are ever in the LA area on Mondays or Fridays and give me some advance notice, I can arrange to attend a SOWG meeting on MER (MSL's are not open to visitors). While some issues are pretty cryptic, some of the decision process will stand out. I think a handful of the SOWG meetings for MER have been recorded and should be available to the public (+). The most difficult part of train new RPs is making them understand the decision process we follow, how to balance risk and benefits. It is a quite involved, often frustrating process.

Regarding driving at night, I have already said that it would be very expensive in terms of power but most importantly in terms of risk. Risk needs to be balanced by a benefit and of all the planetary geologists, including the best experts in this type of terrain none of them have said there's any correlation between temperature and cohesion. None. I have no idea what your background is, but if you feel strongly about this send an email to JPL.

Avoiding sand dunes is not wasting precious time. If I can only drive 20m each night to traverse on sand dunes, while I can traverse 40-60m during the day, I can cover a route that is three times longer and with a much lower risk to the mission. You might add more wear and tear on the wheels but if the rover traverses into sandy terrain that prevents any motion at all (I hate to say "gets stuck") then it would be very hard to do much science (and it would not look good on my resume).

I apologize to the Admins for the lenghty post.

Paolo

(+) a quick google search pointed me to this video: https://www.youtube.com/watch?v=wgyvdf7WJ-I

Posted by: Gerald Aug 30 2014, 01:11 PM

Instead of driving at night one could consider a driving test in the early morning to avoid the VO and AutoNav issues, but the heating issue remains.

Respecting the "electrostatic continuity" I'm with Serpens. The gaps in this idea make it difficult to duplicate the reasoning.
There are lots of possible electrostatic effects which may be considered:
- Currents in conductors and semiconductors are dependent of the electric resistance, which is dependent of the temperature. But most of the sand is insulating. Hence the degree of electrostatic continuity in those settings aren't relevant for the driving.
- Surface charging by ultraviolet due to the http://en.wikipedia.org/wiki/Photoelectric_effect. This may be expected to charge the sun-exposed locations positively and the shadows negatively. There is a difference between day and night. This charging may affect aeolian dust, and the top few micrometers of the surface. Negligible for subsurface and driving mechanics.
- http://www.fp7-spacecast.eu/help/bg_sp.pdf may charge the surface positively, leading to a repulsion within the surface layer. There might be some day/night difference. The effect is tiny in comparison to gravity for the subsurface.
- https://www.mri.psu.edu/faculty/STM/media/wolf-temp-jap.pdf. But it should occur in measurable quantity for a tiny fraction of the sand, at most. Induced electrostatic forces negligible for driving.
- http://www.researchgate.net/publication/249328125_First_observation_of_the_temperature_dependence_of_the_van_der_Waals_Casimir-Polder_interaction_in_the_condition_of_thermal_equilibrium. These are of electrostatic nature in some respect, and may affect the cohesion of particles.

The only relevant chain of evidence between temperature and driving progress I could imagine is temperature-dependent http://link.springer.com/book/10.1007%2F978-3-662-04283-0. Friction is an easy topic, if one doesn't look closer. But it's exceedingly complicated in detail. And it's mostly of electromagnetic nature on the atomic scale.
But there is lot of theoretical and experimental evidence, that among several other factors friction can be temperature-dependent, e.g. for http://scitation.aip.org/content/aip/journal/apl/88/12/10.1063/1.2187575.

Hence your suggestion might be adjustable by replacing "night" by "early morning", and "electrostatic continuity" as the underlying physical principle by "possible temperature-dependent friction".
But I'd suggest this only in the emergency case of stuck wheels, not in normal mode. The effect may well be contrary to what you expected.

Posted by: Actionman Aug 30 2014, 03:13 PM

The dunes are dust rather then sand, the dust forms bonds the can look like sand some of these form linearly.

What about the rovers boom-arm will that be any help ? pushing off in the dune dust?

Posted by: Gerald Aug 30 2014, 03:21 PM

The arm is too precious to risk damage using it that way.

Posted by: SpaceListener Aug 30 2014, 03:58 PM

Curiosity's arm is not designed for land pushing purposes. There are more easy and better solutions than just pushing which is to trace the best route that is always possible.

Posted by: RoverDriver Aug 30 2014, 04:59 PM

No, the dunes and ripples are *covered* by fine particles for only a few mm but the supporting structure is sand.
And here we go again, just like Spirit at Troy EVERYONE was suggesting using the arm to push the rover. Even if we were so foolish to try this:

1) the motor controllers are shared between arm and mobility so you cannot drive AND use the arm

2) the gears in the actuators of the arm cannot withstand the load. Think of it like you in your car stuck in mud trying to push the car with a broom.

Paolo

Posted by: nprev Aug 30 2014, 06:21 PM

Paolo, I'd also add a number 3 (and amplify Gerald's comment thereby): Some of MSL's most critical equipment is contained in the arm, so even if using it as a lever were feasible doing so might well cripple the rover's ability to execute prime mission functions (SAAM).


MOD MODE: I strongly urge all members who are not professional participants in the MSL mission to carefully review Forum rule 2.6. The admin/mod team is closely monitoring the tone of this discussion.

Posted by: elakdawalla Aug 30 2014, 08:30 PM

Paolo, let me speak for all the Admins -- in fact, I think for the whole forum -- that we would never think one of your posts is too 'lengthy'. Thanks very much for your patient and clear explanations; I'm learning from them. It's news to me that you even considered driving in the dark at one time! I thought your drive lengths were most limited by when the communications passes happened. If you did drive in the dark, you wouldn't be able to get post-drive imaging in time to plan movement the next day, would you?

Posted by: RoverDriver Aug 30 2014, 09:15 PM

Emily, it did not go much past the idea of driving after the afternoon comm pass. You are right that the post-drive imaging would need to wait for sunrise, possibly after the morning comm pass which would require more heating for the mast actuators and camera electronics. Really driving at the time we are driving (typically noon-ish to early afternoon) is the most efficient usage of the available resources.

I welcome suggestions and don't want to stifle creativity. I really like interacting with people who might have ideas. Unfortunately one of teh few things we are not at liberty to publicly talk about is anomaly resolution. This is unfortunate because it would really show what issues we are really facing and how my extremely clever colleagues figure out how to solve them. It is quite amazing. There are a few published articles such as the Spirit Sol 18 Anomaly that give a glimpse into this.

Paolo

Posted by: djellison Aug 30 2014, 09:32 PM

There are some great articles on that at the JPL Tech Report Server...

The Spirit anomaly story : http://trs-new.jpl.nasa.gov/dspace/handle/2014/38849
The Mars Rover Spirit FLASH anomaly : http://trs-new.jpl.nasa.gov/dspace/handle/2014/39361
Mars Exploration Rover Spirit Sol 18 anomaly : http://trs-new.jpl.nasa.gov/dspace/handle/2014/40546

Posted by: Tom Tamlyn Aug 31 2014, 12:44 AM

I'd like to second Emily's thanks to Paolo, and also mention that quite a few of the responses on this thread have been interesting and instructive. I particularly liked Gerald on friction ("Friction is an easy topic, if one doesn't look closer. But it's exceedingly complicated in detail. And it's mostly of electromagnetic nature on the atomic scale.").

Posted by: nprev Sep 2 2014, 01:49 AM

Heads up, everyone: I'm gonna move this metal fatigue discussion over to the http://www.unmannedspaceflight.com/index.php?showtopic=7658&st=240&start=240 thread.