Kaivan Karimi – Dir. of Global Strategy, Microcontroller Group, at Freescale

Weekly PULSE Issue

Kaivan Karimi – Dir. of Global Strategy, Microcontroller Group, at Freescale

Featured Engineer

Interview with Chris Vermillion

Chris Vermillion

Chris Vermillion - Altaeros Energies

How did you get into electrical engineering and when did you start?

My journey into the electrical engineering field, specifically control, was initially a roundabout one. My undergraduate studies were in aerospace engineering, and I was specifically fascinated with aerodynamic design. This fascination led me to assume the role of lead aerodynamicist for the University of Michigan Solar Car team, where in addition to designing the aerodynamic shape of the vehicle, I took on a side role of designing the cruise control system. Though it was a simple PI controller, with a few additional bells and whistles, it made me realize that I was pretty good at this fusion of math, physics, engineering, and programming that comprised control. I continued advancing my knowledge, spending some time working on vehicle stability control (where I advanced my programming and embedded control design skills), eventually earning my Ph.D. (and advancing my theoretical and mathematical background), and ultimately taking my current, very exciting position as a lead engineer for Altaeros Energies.

How did you become involved with the AWT?

When I joined the team, Altaeros Energies was approximately one year old, and in the beginning stages of engineering its first functional prototype airborne wind turbine (AWT). It had become apparent that a high degree of autonomy was required to minimize operational costs, and that a high level of stability and control know-how was essential for the operation of a system where the conventional tower was replaced with tethers and a blimp. My introduction to the team had solar car roots, as a former Michigan solar car alum had served as an early-stage mentor to the team and was able to connect the team with other key alumni who, based on their backgrounds, might have been good fits for the Altaeros team. It turns out that Altaeros was especially interested in hiring a lead engineer with a great deal of control design expertise and solid working knowledge of aerodynamic design as well; it was a perfect fit.

As the Lead Engineer, managing areas such as dynamic modeling, embedded control implementation, and sensor selection, what has been the largest challenge?

As a small team (5 full-time engineers) working on a fairly big project, our collective knowledge does not cover every topic necessary to design and build a successful AWT. As a result, it’s been extremely important for us to build up a strong supporting team of advisors, consultants, and engineers to fill these inevitable gaps. Personally, my strengths are in physics, flight dynamics, control algorithm design, and software design. Recognizing these strengths, I’ve continually augmented the Altaeros team with engineers and consultants who are well-versed in hardware design and power electronics. The big challenge here is in management, particularly regarding the level of scrutiny placed upon short-term employees and interns. On one hand, employee work quality and morale suffers under a very high level of scrutiny, and I have to devote an excessive amount of my time toward applying this level of scrutiny to their work. On the other hand, it is essential that I am well-aware of the work of temporary employees so that I can assess how it fits within the big picture of the project and so that I can ensure that continuity is maintained when an employee’s contract with the team is finished. So it’s a tricky balancing act, and one that I am continually trying to hone.

What are the major accomplishments to-date for the AWT?

The Altaeros team recently returned from successfully testing our first functional prototype up in Limestone, Maine, at the former Loring Air Force Base (now decommissioned). Here, we demonstrated power production (which increased by more than a factor of 2 at modest altitudes (60+ meters) vs. ground-level, due to stronger winds aloft), flight to 104m altitude, autonomous take-off and landing, closed-loop attitude and altitude control, and general flight performance properties. Flight performance was demonstrated within a testing period from February-March, 2012, in a variety of wind and (really, really cold and snowy) weather conditions.

What are the key control, sensing and communication components for the AWT?

The AWT control system consists of 3 nodes, namely a host computer, base station, and airborne shell, which are linked via a 2.4GHz wireless network. The host computer is equipped with the GUI with which an operator controls (issues setpoints to) the AWT. The base station is equipped with winches that regulate tether lengths, in addition to a TI C2000 series microcontroller, NI CompactRIO, and a suite of sensors. Finally, in addition to housing the turbine itself, the airborne shell houses its own TI C2000 series microcontroller, along with a sensor suite for airborne attitude, altitude, and weather measurement. The main control algorithm is executed on the base station, whereas the airborne shell microcontroller is used strictly for signal processing and communication.

What is the future for the AWT?

Our functional prototype demonstrated a solid level of closed-loop dynamic performance and autonomy that we are now working to build on for future systems. One development that we’ll be pursuing in the near future is the customization of our turbine power electronics and turbine geometry itself, both of which were based on an off-the-shelf Skystream 3.7 turbine for our initial proof-of-concept. This customization will be accompanied by a scale-up in power production, to at least 30kW on our next system. The next system will fly at altitudes up to 600m, where it will gain access to winds possessing 5-8 times the power density of ground-level winds. Furthermore, we’re looking to achieve extended autonomous operation for weeks at a time with our next system. This will involve aerodynamic refinements, in addition to remote monitoring and control features that will enable the tracking of the system’s performance over a web-based UI, ultimately from a remote base.

What are your favorite hardware tools that you use?

Through a partnership between National Instruments and Greentown Labs (a Boston clean-tech incubator, of which Altaeros is a member company), we has been fortunate to have gained access to a CompactRIO and a whole line of C-Series modules, which have been invaluable for rapid prototyping and processing of a wide array of analog measurements. While our core control system was implemented on a TI C2000 series microcontroller, NI hardware still played a valuable role in aggregating and processing analog measurements on the system and in providing us with a measurement and data acquisition platform for our external meteorological tower at our test site.

What are your favorite software tools that you use?

As a control system designer, MATLAB/Simulink is an invaluable modeling, design, and implementation tool. Through the capabilities of Simulink, Altaeros has been able to construct a 6 degree-of-freedom dynamic model of the AWT. Furthermore, through MATLAB/Simulink Coder (formerly Real-Time Workshop), C-code can be automatically generated from a control system that is designed and tested in Simulink. This provides a powerful model-based design framework wherein a control system designer can build and simulate a controller within the context of the closed-loop system, then implement this controller exactly as it was designed on a supported microcontroller, such as a TI C2000 series microcontroller that Altaeros presently uses. Low-level C programming is then focused not on the core algorithm design but on the signal processing and communication design that is necessary to properly condition the inputs and outputs of the automatically generated portion of the software.

What is on your bookshelf?

Well, with the rise of e-books, electronic journals and conference proceedings, and the general progression of all things written making their way into digital form, I guess the bookshelf is becoming a bit outdated. I wonder what my kids and grandkids will think about these books that people used to use…maybe their thoughts will be similar to my thoughts about my parents’ and grandparents’ record collection. In any case, I still have hard copies of flight dynamics textbooks, controls and programming textbooks from college, and a lot of travel guides from the places that I’ve visited or intend to visit…and I imagine that these will remain a part of my bookshelf for years to come.

Do you have any tricks up your sleeve?

Well, I’m no magician, but I think there are a couple of “tricks” that have enabled me to achieve success in some pretty cool projects. I believe the secret has been knowing what I do not know. As I alluded to in one of my earlier answers, it’s really impossible for a small team (especially a team of 5) to have all the answers when it comes to a project as big as the AWT. But by surrounding ourselves with knowledgeable people and understanding where our strengths and weaknesses were, we were able to accomplish a major milestone.

What advice do you have for students getting started in Engineering?

If you’re an undergraduate, or just graduating, pursue breadth of knowledge and experience as much as possible. To the extent that you can, pursue student projects like solar car or DARPA challenges, and when you graduate, take a serious look at working at a startup. These are experiences that, unlike working for a very large company, will give you the opportunity to play a lead role in an engineering project from start to finish. You will interact with vendors, sponsors, and investors, and you will develop a better understanding of how the pieces fit together in the engineering of a system and the building and sustaining of a business. Furthermore, I believe that a significant level of broad engineering knowledge and experience is essential prior to going on to pursue an advanced degree, which will likely focus on specialization and depth of knowledge. After all, following this advanced degree, you will still need to interact with engineers from many areas and speak their language!

What advice do you have for students studying control?

Become proficient in embedded control design! Take embedded controls classes even if they aren’t required; this knowledge is far more important to the practical implementation of control systems than the theoretical probability class that typically is required at the onset of any graduate program in control (nothing against that probability class; it’s just that the stuff learned there is less important and less frequently used than embedded control design) and is just as important as linear feedback control design theory. Also, pursue student projects and internships that involve embedded control design; this is a great opportunity to practice what was learned in the classroom.

Previous Next

Previous Spotlights

Featured Products

Click Here