26 March 2008

Why does it cost so much?

Posted by Ryan Anderson

We got a question today that I thought I should answer here on the blog so that others who are wondering the same thing can have an answer. The question was:

Since the rover is already on Mars and it’s solar powered why does it cost $4 million a year to operate? I’m not doubting it does I just don’t know how projects like this work. Is most of it tied up in the salaries of the scientists working on the project, the computer systems used to analyze the data, or what?

It’s a valid question. Actually, it costs $20 million per year total to run both rovers. The $4 million is the amount that was going to be cut from what remains of this year’s budget. Now, to most people, $20 million is a lot of money! I know I would be hard pressed to spend that much money in a year even if I was trying. So what gives? Why do two robots on Mars cost so much to maintain?

The answer: robots don’t need to be paid, but people do.

There are hundreds of scientists and engineers on the MER project. Some of them, like myself, work on the rovers only occasionally, maybe one week every two months. Others work long hours every day doing nothing but keeping the rovers healthy and doing good science. It may help if I describe a day in the life of the rover team to give an idea of what is involved.

The day begins in the middle of the night: Mars Odyssey relays the data that the rover has collected on the previous day to the so-called “Deep Space Network” of radio dishes on Earth, who receive the data and send it to the NASA Jet Propulsion Laboratory. From there, the data is sent out across the world to the many universities and research institutions where there are scientists involved on the project.

In the morning, the scientists take a look at the data and see what it shows. Maybe the rover was analyzing an interesting rock and returned the results, or maybe it was taking a bunch of color pictures, or monitoring the amount of dust in the atmosphere. Often, it is a combination of these things. The scientists make sure that their instruments are healthy and in working order and that they received the data they were expecting. Then we think about what comes next. Do we drill into the rock we’ve been looking at to see what it’s made of on the inside? Or do we want to drive close to that crater in the distance?

Meanwhile, while the scientists are checking the data that came down, engineers at JPL are making sure that the rovers are safe: were the temperatures high enough for the computers to run properly? what was the solar array power like yesterday? did we drive and end up where we expected?

Around midday eastern time (morning at JPL in California), the science team meets and discusses what data we’ve received and what we want to do next. We come up with a plan and pass it on to the engineers. The engineers take a look at what the scientists want to do (You want to drive on how steep a slope?!) and then come up with a way to achieve that goal. Then, the engineers have to convert everything that we want to do into commands that the rover understands. Finally, those commands are beamed back to Mars, where the rover receives them in time to follow them for another day.

And remember, there are two rovers.

We do what we can to streamline this process. Often we plan two days at a time, and we plan three days on Friday to get through the weekend without having to meet. After more than 1500 days of operations on this 90 day mission, the team is as efficient as it can be. But it still involves many hard-working people.

All those people are passionate about space exploration, and love what they do, but no matter how passionate we are, that doesn’t pay the bills. People need to be paid for the work that they do. When you have hundreds of people, that cost adds up very quickly. Add on to that the cost of maintaining up to date software that we use to plan what the rovers do, and updating the software on the rovers themselves (their original programs were never meant to run this long, so things like dates and times need to be updated occasionally). It’s not hard to get up to $20 million.

To put all of this in perspective, $20 million is really not a lot of money on the scale of government agencies. It is about 0.15% of NASA’s $16 billion budget. And NASA’s budget is about 0.7% of the US federal budget. Put another way, the United States spends $20 million every 4 minutes or so, and $20 million keeps the rovers running for a year.

So, it costs a lot, but there’s a good reason for it, and it’s a drop in the bucket compared to the national budget.