Monday, August 20, 2012

Why don't they just..?

It seems every time I come across a story about the Mars Curiosity rover there will be many people commenting on the technology used starting with "Why don't they just..?" and usually pointing out things like: the processor in their smart phone is way faster than the one of Mars, or they have way more memory on their iPad, or their digital camera is way better than the one sending back pictures. These "Why don't they just..?" questions are both annoying and to be expected.

Annoying because the underlying thought is "Those NASA/JPL guys are so dumb LOL" and to be expected and encouraged because we wouldn't make any progress without asking questions and, in particular, asking why.

But it doesn't take much research to find the answer. (Even though I'm tempted to answer: "Because it's on friggin' Mars, doofus!")

1. The Mars Science Laboratory project was started eight years ago in 2004. So, all the technology on it is at least eight years old.

2. The trip to Mars means flying in an area with high amount of radiation (from things like cosmic rays, all manner of stuff flying out of the Sun and the Van Allen radiation belt). That means all the electronics needs to be radiation hardened. So, you don't start with just whatever you can get from Fry's Electronics in 2004. You need specifically radiation hardened components like the RAD750 processor in Curiosity.

3. You need to be a bit conservative. The thing you're sending to Mars is going to be on its own and unrepairable. It had better work. So, you're likely to reuse components and techniques that you know work. It has been reported that the skycrane used to land Curiosity used components derived from the 1970s Viking landers and algorithms used on the Apollo craft.

4. And once you've worried about radiation hardening, reliability, and weight you need to worry about bandwidth back to earth. It's no good taking gigapixel photographs if you can't get them back to Earth. For example, Curiosity can communicate with the Mars Reconnaissance Orbiter for a few minutes per day at 2Mbps and with Mars Odyssey at 256kbps.

But rather than explaining all this stuff, I think there's a better way: build, land and operate a rover here on Earth.

The Rover Challenge

I've done one high-altitude balloon flight and watched the progress of many others. Although going to Mars is a very different situation there are similar challenges: weight, environment, communications, landing.

A good way to see how hard it is to build and operate a rover would be to build one designed for operation in an inhospitable part of Earth. Launch it via a high-altitude balloon with parachute descent and then operate it without GPS over a slow, high latency radio link.

It would actually be a fun project. On a balloon you can probably have about 2kg of payload maximum for your rover. Now imagine 2kg weight budget for a semi-autonomous rover that would be dropped into a desert in the South Western USA, or the Sahara, or the Australian outback.

The rover would have to withstand high tempertures, dust, and wind; operate on a perhaps unstable sandy ground; communicate using HF radio; and operate without a human touching it.

To me it sounds like a fun challenge. Anyone else?

If you enjoyed this blog post, you might enjoy my travel book for people interested in science and technology: The Geek Atlas. Signed copies of The Geek Atlas are available.

24 Comments:

Blogger Bartosz Banachewicz said...

Sending a rover to sahara by a balloon (or even better, a zeppelin) is by far the most wicked yet doable projects I've ever seen.

9:45 AM  
OpenID eroomde said...

I've always quite wanted to land a little rover into some kind of toxic and human-unfriendly semi-active volcano mouth.

I assume bonus points for retrorocket-assisted descent?

10:03 AM  
Blogger carbonfeed said...

+1 for into the mouth of a volcano, if it's going to be a one way trip, make it a great one!

10:39 AM  
Blogger wisdomdata said...

http://www.travelling-australia.info/CollPaintdes/APics3B.html

Somewhere like this near Coober Pedy - its like mars with trees and retrieval could be arranged.

1:18 PM  
Blogger Thomas Crescenzi said...

I would absolutely be up for devoting time and resources to such a project. The volcano idea would be pretty damn absurd.

1:36 PM  
Blogger Thomas Crescenzi said...

This comment has been removed by the author.

1:36 PM  
Blogger Chris McClelland said...

What happens if it happens to land in a really inhospitable environment, like, say, a South London council estate?

2:14 PM  
Blogger Chris McClelland said...

What happens if it happens to land in a really inhospitable environment, like, say, a South London council estate?

2:15 PM  
OpenID catch22 said...

So why did they push a software update remotely, and not just switch to it remotely?

4:35 PM  
Blogger Kerry Wilson said...

That does sound like a fun idea.

I assume there is also limited power. I don't think we want to take 8 years to build a robot just to have it last a month b/c the geek in us wanted to put the latest and greatest multicore processor in it.

5:07 PM  
Blogger Ricardo Tomasi said...

@catch22 according to interviews, limited storage, and they were still working on it while the rover was on the way to mars.

6:20 PM  
Blogger Henry de Boisseguin said...

great great great idea there. I can foresee the heyday of Arduino.

9:08 AM  
Blogger Henk said...

I'm doing robotic project right now, and weight - is not only restriction. Also - power that your battery could generate. And that power restricted by weight too. So if you want to have 50Ah, with battery weight ~ 100-200gr - it is just impossible.

7:51 AM  
Blogger Brian Kuehn said...

I'm actually closer to the arctic circle... That is a pretty inhospitable place!

9:39 AM  
Blogger jenelia desouza said...

This comment has been removed by a blog administrator.

10:25 AM  
Blogger Peter Wallhead said...

The Rover Challenge sounds awesome! I'm asking around my friends now to see if something like this is doable.

10:37 AM  
Blogger Matthew Hall said...

Curiosity does not even have NFC ;-)

11:31 AM  
Blogger Traveller said...

What about funding it via Kick Starter as an open source learning project

12:39 PM  
Blogger tbohon said...

One of the neatest ideas I've seen floated yet. As someone with a scientific curiosity and background in computer simulation and also as a ham radio operator I'm virtually salivating here ... :)

4:54 PM  
Blogger Fabrice Leal said...

OMG, thats the coolest think I ever thought of! A journey to the arctic circle ...

6:39 PM  
Blogger mr_geronimo said...

i'd send mine to Detroit..

7:27 PM  
Blogger Dan Frederiksen said...

John, when a single geek does what you propose on a shoestring budget with much higher image quality and even bandwidth despite not having access to a satellite network like nasa could have done, will you then wake up and realize how pathetic nasa is?
what Curiosity is doing would not be out of place in 1970. it is that pathetic. indeed it could be outdone with live video in 1970

11:21 PM  
Blogger A_flj_ said...

My approach would be different: build and launch an entire horde of small, inexpensive robots which can fix each other, scavenge on each other for parts, and collaborate on various tasks.

Sort of deploying a mobile cloud, instead of a single powerful server, in terms of software.

It would be a real challenge for the software developers, but not really that much of a challenge for the hardware developers, IMO.

Comming to think of it, the rasp.pi I just received plus a lego mindstorm make a good prototyping platform for prototyping ...

8:37 AM  
Blogger Keith Chuvala said...

> Dan Frederiksen said...

"...when a single geek does what you propose on a shoestring budget...."

Pray tell, who, and what project, are you referring to?

Full disclosure: I've worked for a NASA contractor for over 13 years. I have observed very little evidence of "pathetic" over that time.

Well, unless you're talking about political appointees at the top end of the agency, I guess....

11:18 PM  

Post a Comment

Links to this post:

Create a Link

<< Home