Dr. Julie Walker is a navigation engineer at Intuitive in Sunnyvale, California.  After studying mechanical engineering at Rice University, she received her PhD at Stanford University in the Collaborative Haptics and Robotics in Medicine Lab (CHARM). Her research focused on human-robot interactions through the sense of touch, specifically with holdable and wearable haptic devices for medical applications.

I had the pleasure of interviewing Julie, who is based in California, while I was across the globe  in the Netherlands for my studies. We talked about her experience as a university student growing up with two engineer parents, her past and current projects, and her work life at Intuitive.

Although Julie majored in mechanical engineering, she wasn’t interested in medical robotics until she started her undergraduate research internship at the University of Pennsylvania. She refers to that experience as the most impactful thing she did because it made her realize what she could do beyond the required classes. Getting involved in research and seeing the applications showed her how exciting this field could be. She suggests that being involved in research as early as possible can be a huge advantage for students who consider getting a PhD degree.

As a navigation engineer at a medical robot developing firm, she is currently working on a flexible snake robot developed to improve the lung biopsy. Her responsibility is using all the sensors in the robot to help the user navigate to the right target as easily as possible. Majority of her work involves writing algorithms that integrate different data sources into a useful map for the surgeon to utilize. She is programming in C++ and implementing different prototypes to add to the system to help the user understand better where they are in the body and give them directions to get to the right place. Even though she is not using reinforcement learning or artificial intelligence (AI) in her work, her background in data-driven learning and understanding  of the different tools help her decide the best way to guide the user to the correct part of the lung.

Having two engineers as parents, being involved in research throughout her entire career, and having experience in the haptics lab were some of the things Julie says prepared her for her current job. During her PhD, she spent almost a year working in a lab in France, where she got to see another side of mechanical engineering, the medical devices. This opportunity led her to spend more time on medical robotics. Getting a broad view of the different parts of robots was the most satisfying part of the research for Julie.  After finishing her PhD, Julie took a position at Intuitive, which is primarily a software engineering role. Even though her background is in mechanical engineering, she is no longer building any physical devices. However, her background and understanding of the sensors, mechanical parts, and the system as a whole have given her the perspective she needs to advance her career. 

Highlights from the interview:

My first question is about your background. What were your parents’ expectations for you and how did this shape you into choosing your career?

My parents are both engineers, and I grew up knowing about that career path in general. They encouraged me to be more interested in science, math, and engineering. I chose mechanical engineering in undergrad because I really liked that you could see what is going on and understand it physically, whereas electrical engineering to me was harder to conceptualize. Before I started undergrad, I didn’t realize how powerful a computer science degree could have been at the time, but as a mechanical engineering student, I did see the value. I was much more interested in the robotics side of classes, rather than heat transfer, thermodynamics, and fluid mechanics. 

Did you have a mentor who helped you to cultivate your interest in robotics?

After my second year in undergrad at Rice University, I applied for an internship program that matched me with a professor at the University of Pennsylvania. She studied haptic interfaces in medical robotics. She was very encouraging and introduced me to the world of robotics. When I returned from that internship back to my university, I found a professor doing similar work, and I worked with her for the rest of my time in undergrad. Both professors introduced me to the field of medical robotics and human robotic interfaces. My advisor in undergrad, Dr. Marcia O’Malley, encouraged me to consider going to graduate school. So, I went to Stanford to do my PhD and continued working in haptics and robotics, specifically focusing on haptic interfaces for medical training. Can we use the sense of touch to help train robotic surgery systems and teach them when they’re doing the wrong motion and help correct them? What kinds of devices would be best suited for that? What kind of controllers can be used to apply those forces? I started experimenting in artificial intelligence and machine learning and discovering how I can use the data on how a person is moving to train my haptic device to do a better job. 

Did you also write robotics software?

Yes. I both designed the actual physical robots that a person would interact with and wrote the software to control it. So, I had a broad view of all the different parts of a robotic system and how they need to communicate with each other and work effectively. The software was the most satisfying part for me. After finishing my PhD, I took a position with Intuitive, primarily a software engineering role. I’m no longer building any physical devices, even though my background is in mechanical engineering. My background on building sensors, mechanical parts, and the entire robotic system as a PhD student has given me a good perspective for my current role in understanding what else is going on in the larger robotic system that we work on. 

Can you say your consistent and productive background helped you get to where you are right now?

I haven’t meandered very much, I started working in medical devices, and here I am now. I didn’t have any forays into other areas. But, at Stanford, I tried to expose myself to as many different things as I could. I took many classes on reinforcement learning, control systems, mechanical design, and medical robotics. It’s been a direct path for me.

Can you give a brief example of the types of projects you work on right now?

The product that I work on at Intuitive is a robot for performing biopsies in the lungs. It’s a flexible snake robot that is extended down the throat and into the lungs to take biopsies of suspicious looking lesions. We can see the branches in the lungs, but it’s very difficult to drive to the actual spot. We really want to make sure that the user gets to the correct place to take a biopsy sample. My goal is to help the user navigate to the right target as easily as possible using all the sensors in the robot. It entails writing algorithms that combine these different data sources into a useful map for the surgeon to drive in. How do we take all these different information sources and tell the user exactly where it’s driven to and get to the right place? I mostly program in C++ and implement different prototypes to add on the system for the user to help them better understand where they are in the body and give them directions to get to the right place. I am not using reinforcement learning or AI in my position, but understanding the different data-driven learning and planning tools that are available to guide someone to a part of the lung is helpful background. 

Can you give us more information on the existing opportunities in the field of medical robotics?

The field of medical robotics is a hugely satisfying field to work in. It’s also very challenging. Robots can perform tasks in a well-organized rigid environment pretty easily, but when you put them in a squishy body that is changing and moving and is delicate, performing the tasks gets difficult. I think there’s a lot of opportunity in medicine to push the boundaries on what robots are capable of doing. I particularly like this field because it’s not an autonomous robot doing a procedure, it’s a human controlling the robot, but we have a lot of intelligence between the human control and the robotic output. There’s a lot of opportunity to use tools from AI, but in a medical system, anything that is probabilistic is a little bit risky. So, if you can write an algorithm that will to do the same thing every time, that is often a safer choice. 

What was your most memorable experience that helped you develop as a person?

My first research experience in undergrad that exposed me to robotics was the most impactful thing that I did. Getting involved in research and seeing the applications showed me how exciting research could be. 

What keeps you motivated at work?

Trying to treat my job as a learning experience is a healthy attitude for me to understand that every time I’m struggling with something, it’s because I’m learning something new. That is why I wanted to take this job in the first place because I knew that it would give me the opportunity to learn new skills. I am spending a lot of time trying to get better at writing codes that are easy to read, more stable, and more useful in many applications. I get feedback from my teammates to improve; it sometimes doesn’t feel good to have a lot of feedback, but you don’t get better by doing things correctly, you get better by doing things wrong and learning from it. Knowing that I could be writing a feature that is going to make it easier for a physician to make sure that they hit a cancerous lesion is very rewarding. Hopefully, doing biopsy and treating cancer in one procedure will be very impactful. 

Can you describe the team dynamics in your project?

Half of the people on my team  are computer scientists, who focus on the structure of the code and the GUI. The other half focus mainly on algorithm design. I spend a lot of time brainstorming with other people trying out different prototypes and looking at data. I spend 60–70% of my time on a long-term investigation project and 30% on day-to-day, shorter term changes on our robot from surgeons’ feedback. I also work with user experience designers as well as clinical design engineers, people who train the surgeons and interact with hospital representatives to understand the clinical needs. 

What are the changes or challenges that you foresee in this specific area?

There’s always a challenge with user facing devices; people feel differently about the way the system works. Some doctors do not like the way we implement certain things. For some doctors, something would be really intuitive, but for another doctor, it would be very confusing. I’m trying to make sure that we’re getting enough data on how people feel and how easy it is for them to use this system, their performances, and make sure that we are not just designing for one class of people that will be interacting with the system. We want all surgeons to easily use the system to get to the lesions.   

What advice would you give a student who is interested in your field?

They should get involved in research early on. As they work on the research projects, they should not only think about what they’re doing but also look at what everyone else is doing. Especially in the beginning, in a project that you might start early on, you don’t have enough skills to do something complicated, but look at what the people above you are working on and think about whether you would also want to do those things in the future. Because even if your project is small and maybe you don’t feel engaged by it, it’s such a great opportunity to learn new skills. Talk with the other people in your lab about what they imagine doing with their degree, how they got there, what skills they use that you should work on. People like to talk about themselves; so, it’s easy to ask people questions about their jobs or their background and you can learn a lot about different careers.

Interview excerpts have been lightly edited for clarity and readability and approved by the interviewee. This article only aims to share personal opinions and learnings and does not constitute the interviewee’s current or former employer(s)’ position on any of the topics discussed.