Charles "Chick" Macal

By Louise LernerFebruary 5, 2012

Charles (Chick) Macal is a senior systems engineer at Argonne National Laboratory. He specializes in a type of computer simulation called agent-based modeling, which calculates likely decisions for each "actor"— one person, for example—in a simulation and then sees what impact those have on other agents. He has applied agent-based modeling to simulate the spread of MRSA bacteria, to forecast economic panics and to predict consumer behavior, among other projects.

How did you get interested in science?
I think when I was young I was fascinated by the natural world. I actually started out collecting butterflies and insects when I was under 10 years old. My parents took me on many trips to the Field Museum in Chicago, and I loved dinosaurs and things like that—but then it developed into this interest that didn't go away.

When we landed on the moon, I was just totally enthralled by every single step, even from the beginning, with John Glenn. During the moon landing my teacher would send me down to the TV room during class and have me report back on what happened—we only had a TV in one room in the school back then.

Then when I was in high school, I found that science, and particularly mathematics, were the things that I was most interested in. I actually read a book in the library about what engineers do, and I got really intrigued by engineering; so I went to Purdue and studied engineering. There I got interested in applying engineering to new kinds of problems.

Why engineering?
There are so many problems in the world that need to be solved; and so many of them have to do with large-scale systems. We live in a society full of technology, and systems of technology—everything from transportation to energy and its production and distribution.

When I came to Argonne in 1975, it was the midst of the energy crisis. There was no higher calling, in terms of solving problems of national significance, than solving the problem of energy. Argonne was a tremendously exciting place to come and do that kind of work. So, you know, I'm still working on that [laughs]. We're looking at things like new markets for solar photovoltaics, and the process by which the market for solar grows, and what factors help that along. It's all about innovation; how it occurs, and how it penetrates the market. That's part of the R&D challenge—to understand how to move discoveries from the lab to the market. I like to look at the system as a whole and how the parts fit together, not just the individual parts.

What are you working on right now?
There are three main projects I'm working on; modeling the spread of MRSA, modeling solar markets that I mentioned earlier, and also a project funded by the Department of Defense, looking at human social, cultural and behavioral factors, and how those systems respond to military actions. One of the things we're looking at is taking theory from political science and sociology related to conflict—how political movements evolve and grow, for example—and how those theories can be quantified and put into a computer model, so that we can simulate when and if conflicts may arise.

And you can quantify that?
Well, that's what we're working on. We're still doing quite a bit of research. What we do is take a country and look at its history to essentially calibrate a model: understand the movements and what situations led to revolutions or other political conflicts. Then we see if we can make that model useful elsewhere. But rather than just having indicators—like poverty, income levels, or disparity—which are statistically based, this goes into the causal mechanisms a little more deeply. That is, how people respond, in a broad sense, to rewards and punishments that movements might use to try to influence the population; as well as the social interactions that occur between individuals that turn them toward one way or the other.

With so many very different projects, what field would you say you're in?
I think the field I'm in might best be called "computational social science." We partner with people who are experts in that field, so that we can get depth. What we bring to the table is the modeling capability and experience—when I say modeling, I mean translating mental or qualitative models into mathematical algorithms, and eventually computer code.

What's the biggest challenge in your field right now?
I think the actual biggest challenge is to understand how to model human behavior, in a way that is credible and conducive to allowing us to build models that we can validate using tests. I will say that it's not actually as big a problem as it might sound, because we're not trying to predict all possible behaviors; just behaviors in a fairly narrow, specific situation.

But it's hard because there hasn't been much work done on how to actually model behaviors. There's been a lot of work on surveys and describing behaviors, but not in terms of causal models. But if we had good models, of how people decide to do things, we could do really wonderful large-scale models with millions of "agents."

What's one new project you're looking at?
We have been approached by the Federal Highway Administration. They want models of traffic that have better models of behavior for the drivers. It's both driving and route planning—why people take which roads—and also things like gaper's block, where drivers slow down to look at an accident.

There are very interesting possibilities concerning information: if drivers had real-time information about the traffic situation, which is likely in a few years, what would they do? If they all saw the expressway blocked ahead and then all got off at the same exit, what would happen then?

Traffic models right now are more physics-based: drivers as particles flowing down lanes. But each of those "drivers" doesn't have an individual mind; each driver just does what every other particle does in the model.

Do your models account for people who buck the trend?
Our models definitively do account for some who don't, or can't, follow instructions. Especially in, say, an evacuation model, the issue of how people behave in emergency situations is very complex. People exhibit quite a range of behaviors in a crisis.

What keeps you interested in your work?
What keeps my work enjoyable is that I'm constantly coming across problems that have not only not been solved, but the questions haven't even been asked yet. That's very exciting. So we quickly get to the forefront and often we can even make a contribution.

There are always things in the news that are somehow related to the work that I'm doing. It's all totally relevant. Everything I'm doing is relevant to today's world, and also the future. I never know what next week will bring in terms of possibilities. I really try to focus on the bigger problems of national importance—the problems of the world that really need to be solved.

There are those who say that making computer models of social and behavioral processes cannot be done. But we are working in areas and making models of processes that no one has attempted before, so it's not clear what cannot be done. I often feel that a grand discovery is right around the corner. We just need to be in a position to recognize it for what it is when we come upon it.

Did you have a role model?
In college I had a real role model. One of my professors—who was also my advisor: A. Alan B. Pritsker—he could look at a situation or a meeting in which everyone was in disarray, and he could sort out the situation right there on the spot, and create a path forward that everyone could agree on. He was a great inspiration. He was the person about who I said "I want to be just like him." If you want to be interdisciplinary, to work with people from many fields, you need that skill. And interdisciplinary work can really, at its best, generate an immensely creative synergy. I've seen it happen many times over the years. There's a lot of energy that's created in these processes that I can draw on.

What do you do for fun in your free time?
I'm on the editorial board for a couple of journals. That keeps me busy with reviewing articles, but I feel that I am contributing to help advance the field of agent-based modeling. I like to go running, too. It's nice to get away and to think. I've had some of my best ideas while jogging.

I also like to ski moguls and hike in Colorado on vacation, and I teach a course at the University of Chicago's Graham School of General Studies in the Threat and Response Management department. I love teaching that because such a wide range of people come through my classes: doctors, lawyers, police and firemen, public relations folks, first responders. Each brings a unique perspective to the class.

Dan Abraham »