It's probably been bugging Fiance S. a bit, but, I've been enthusiastically following the progress of the NASA Spirit and Opportunity rovers on Mars. The reasons behind my enthusiasm are threefold: I actually interned at NASA for a summer back in the mid-90's and worked in a lab next to a team who did locomotion software for planetary probes like this, one of my best childhood friends is an astrophysicist at JPL where the probes are made, and for a kid who wanted to be an astronaut as a child, this is very cool stuff.
I also had to feel a certain amount of sympathy for the controllers of the Spirit rover when it malfunctioned last week. One of the things that kept a paycheck coming for me during my years in graduate school was a job for a project that collected weather and ocean data funded by the U.S. Office of Naval Research. Part of this job involved keeping data flowing from a variety of instruments into a central database.
Dealing with problems at instrument sites was always a pain. The instruments were often a good many miles away, and, stuck in locations that were difficult to get to. We used to access them via a network of radios (similar in concept and technology to the wireless ethernet available for homes now) that were usually pushed to their limits and a bit flaky. If we were lucky, there would also be some kind of leased phone line to a nearby site because someone else was also interested in instruments located there.
Problems would happen a lot, too. The equipment sat in locations that weren't ideally hospitable for computers. Sheds would warm up to over 100 degrees F during the day. Salt air had a tendency to corrode everything. Storms would knock down antenna poles. Some sites had flaky power connections and used backup batteries that would leak. I can remember being warned not to kneel on the floor of one particular shed where we kept equipment; battery acid was slowly eating through the wood floor and would ruin a pair of pants by the time we got home if we knelt on it.
So I had to spend more than one afternoon trying desperately to communicate with a computer sitting out in a shed on some isolated rocky beach or near some desolate rock trying to restore the flow of data from there to the project's database. The computers were often several years old, and the methods communicate with them (radio or modem) were frequently slow as molasses. Failure would mean "a site visit". Instruments were often originally set up by someone at a partner institution, and, a site visit would entail scheduling a time when I could go with someone from that institution to visit the sick computer in question.
The people driving the Mars rovers don't even have the luxury of a site visit, uncomfortable as it often was. I can only imagine what it must be like to have something go wrong with equipment that is 100 million miles away on a nearby inhospitable planet. The sense of panic must have been especially pronounced, given the number of Mars probes that have just disappeared over the years. The sigh of relief when Spirit finally did check in last week must have been equally palpable.
I hope that nothing else goes wrong.
on 2004-01-26 at 1:17 p.m.
The Wayback Machine - To Infinity And Beyond