>> Thursday, July 22, 2010
One of the downsides to having multiple blogs all on the same system is that I sometimes put my post on the wrong blog. This was first put on Rocket Scientist (where The Mother commented: "Two words: R and D." It should have obviously been put here. Sorry about the confusion.
Jeff King asked: The mars rovers look simple to build, so why do they cost so much to build… actually I thought they would look cool and very complex?
That's actually an excellent question. I think a lot of people wonder how space equipment comes to be so very very expensive. And it is. Even simple straightforward equipment is often an order of magnitude more expensive than it would be in a retail outlet.
There are several reasons for this, but those reasons are driven by two other reasons: extreme environment and reliability. The first reason drives design and it's easy to underestimate the impact of environment on hardware.
To rove around on Mars, a vehicle has to be able to withstand dust and unknown terrain.
It must be able to withstand a radiation environment that is far harsher than we have on Earth (thanks to the protection from our magnetic field), which is hell on electronic equipment, computers and data storage. Most electronic hardware made to withstand high radiation environments are created for military and frequently go through extensive testing and/or "rad-hardening" that exposes them to radiation, testing and often redesign so that they can reliably withstand more extreme radiation environments. This not only increases development and testing time, it might increases material costs if different materials are required to reach an effective design. That increases costs enormously, especially given that there are often only a handful of vendors for a part - which means they can charge nearly anything they want for those consumers who must have them.
The electronics and batteries and all other equipment must also be able to withstand temperature extremes, particularly cold which is very hard on many aspects, including lubricants, batteries, electronics. The extreme in cold can affect tolerances for mechanical interfaces. Which means all equipment must be tested in environments colder than most electronics on this planet need to survive for months if not longer. Vacuum (or, rather, the very very low pressure) adversely affect lubricants as well and complicate thermal characteristics as convection effects are limited. That means you might need some form of heating and cooling for electronics that are cold-sensitive or heat-producing.
For reliability, there are design considerations as well. The design has to be restartable and redundant, in case something goes wrong, or the cost of the whole mission (hundreds of millions, mostly on the trip to/from) is lost. Batteries must be lightweight, high density, resilient to radiation and cold, reliable, and long lasting. Solar arrays must also be lightweight, efficient, reliable and able to use the reduced light available on Mars (compared to here).
Because so much hinges on the functionality of these parts, each part must have a "pedigree" - i.e. been obtained through vendors that have certified processes and materials and who ensure that each part is tracked and verified to be clean and functional. And that makes those parts more expensive.
But, what frequently costs so much is verification and analysis. Hardware is tested, exhaustively, through dedicated use (often months) in testing facilities. Often flight-like prototypes are destructively tested to ensure the flight hardware meets requirements. In addition, a great deal of time and manpower is spent performing reliability and safety analyses, fracture control analyses, fault trees and failure mode and effects analysis, safety assessments.
Frequently, complex (and it can look simple without being simple) devices are one-of-a kind and require exhaustive and unique software that must also be exhaustively tested and retested.
No one's going to sue us if a rover fails, but years of design and planning, years of flying through space, and, literally, hundreds of millions of dollars are at stake, not to mention quite a few reputations.
If it's gotta work, you gotta do it right the first time and that costs a lot of money. But it's cheaper than doing it over. Better more money spent on testing than a second flight because the first one failed.
And that's why they cost so much.
And kudos, I think, have to be offered to the teams that built both Opportunity and Spirit for surviving >20X their expected lifetimes. That design overkill meant we got our money's worth.