You are here
Ellen MacLean: Tracking Success
Senior Principal Engineer, Defense Industry
Greater San Diego, CA
Graduation Year: 1986
A senior principal engineer at Orincon, a San Diego company later acquired by Lockheed Martin, Ellen MacLean builds algorithms for tracking and data-fusion technology used by the U.S. military, which enables sophisticated surveillance of areas from deserts to dense urban areas. Much of her work is for the U.S. Defense Advanced Research Projects agency, or DARPA, which means that much of her work is classified. But MacLean freely shares her passion for what she does, and her fascination for bringing these intricate formulas into action.
At Duke I got a broad look at what I could do, from building hardware in EE labs to solving system problems to learning about different algorithms and algorithm development. Duke let me see for myself where I wanted to go.
Tell us about what you do.
Essentially, we build technology that can use geographic location data, taken over time, to come up with an operating picture of all the motion in a particular area.
For the last 20 years our work has been done using multiple hypothesis tracking, or MHT. Data from the area is gathered using a GTMI, or ground moving target indicator radar, and more recently we’ve also been using video sensors developed by DARPA. Then the MHT algorithm we develop sorts through all the data and how data relate to each other, and it develops the most probable models for what’s happening in these different areas of interest. We’re also working with an emerging technique that uses a graph-based approach to represent the data associations; it uses the same underlying technology as our MHT work, but graph-based approaches can be more useful when you’re in a dense urban environment, as opposed to when you’re out in a desert somewhere.
It’s very much a niche technology, and it’s got a lot of applications in the defense arena. You could also apply it to homeland security problems, such as border crossing issues—it can be applied to any situation where you’d want to know about movement in a particular area.
What do you enjoy most about your job?
I enjoy the technical challenge of trying to solve the latest problem. The teamwork, working together in the lab and brainstorming to try to solve a new problem, is an absolutely fascinating process. I also really enjoy going out into the field when we take our solution into live tests. You do all this up-front work and then in real time it never works the same way your simulation did, so you’re having to do a lot of thinking on the fly. I like that combination of being able to be in the lab and then going out and putting our algorithms and software to the test, and figuring out where they’re breaking, and how to fix them on your feet.
How did you get into this field?
It was partly pure luck. When I was at Duke I got a summer internship in DC with a company that did a lot of underwater acoustic work—trying to find submarines underwater. They worked with interpreting sound signals; in other words, when sound passes through water and bounces off something, is it bouncing off the ocean floor or something else? After graduation I went back to work with that company and got involved with folks who were doing tracking work—taking that signal detection and putting it together over time to find out where the object was moving. I’ve stuck with the tracking and data fusion area ever since. In my career, I’ve applied it to underwater acoustics, ground tracking problems, air tracking problems and a combination of surface, ground and air-tracking problems.
How did Duke affect your career path?
Duke set me in the exact right direction to be able to make a career for myself. Even though I applied to the School of Engineering, at the time I thought that maybe I actually wanted to do math. But through the classes I took at Duke, I found that I was much more of an engineer with a mathematical bent. I got this broad look at what I could do, from building hardware in EE labs to solving system problems to learning about different algorithms and algorithm development. I got to explore what I wanted to do and how I could take what I learned from Duke, go to graduate school, apply it there, and then apply it to the working world. Duke let me see for myself where I wanted to go.