Year in Review: 2021

31 Dec 2021
by Mark
Share

Year in Review: 2021

Summary: A copy of my end of year message to my lab, where I review the key events from 2021.

During this second year of the pandemic, we have grown much compared to last year, and we have much to celebrate given the circumstances. As is always the case, the best part of this year has been working and growing with all of you, and I feel perpetually blessed that I am surrounded by such an adaptable and hard-working group of people. I know I say this perhaps too often, but it bares repeating: working with all of you is really the best part of my job and what makes it so special for me. We’ve had a variety of positive news to celebrate, which I will try to highlight below for you, as well as some things to look forward to in 2022.

The People

As I mentioned above, the best part of the lab for me is all the wonderful people I get to collaborate with and call our collective lab family. As an added bonus this year, the arrival of vaccines meant that we were able to actually meet in person more frequently this year, leverage more on-campus resources, and also finally do a lab retreat (Disc Golf!) in person in August.

This year we continued to grow with the addition of three new lab members, Arthur Drake (coming from UMD), Milad Habibi (coming from Northern Arizona Univ.), and Alec Van Slooten (coming from UConn), who are continuing some of our NIH and ARPA-E work on low-dimensional design manifolds, optimization, and Inverse Design, as well as exploring new directions. It has been great to have their insight and energy onboard the team to help out with the work the rest of the team spearheaded last year. Further, we added two “honorary members” to the lab after the birth of Jun and his wife’s twin boys, Jasper and Felix, earlier this year! It was great to see and support Jun and his growing family over Zoom after he came back from his paternity leave.

At the same time, we had a notable departure of Post-Doc Dr. Fatemeh Ghoreishi, who secured and started a position as an Assistant Professor at Northeastern University. We wish her the best of luck in starting up her group there and look forward to seeing her again at future conferences and other events, and meeting her future students. As part of the Maryland Robotics Center Post-Doctoral Fellows program, Dr. Ghoreishi spreadheaded some of the optimization and design work for the soft robotic catheters we were exploring with Dr. Sochol and Dr. Krieger.

The Science

The lab’s scientific work this year – or at least the published work that I can talk about publicly – largely reflected some of the outcomes of projects from last year, since it takes a little while for papers to get reviewed and published. Specifically, some of our work from our past and current ONR, NIH, DARPA, ARPA-E, and NSF projects came out this year. You can get a full list of these by checking out our papers page or Google Scholar, but I’ll highlight a few key threads here:

  • Some of our work from the DARPA Fundamental Design program was published and presented as part of the ASME IDETC 2021 this year. This included Nicholas’ paper on Automatically Discovering Mechanical Functions From Physical Behaviors via Clustering, which tries to find new and unknown mechanical functions given only examples of physics simulations, and is an attempt to automate mechanical discovery. Also, Charlie presented his paper on Potential Energy Surfaces for Conceptual Design and Analysis of Mechanical Systems which presents a new abstraction for designing and reasoning about multi-physics mechanisms that significantly extends the usefulness of traditional Configuration Space (or C-Space) methods for Mechanical Design. I thought it was a very thought-provoking piece of work that helped me fundamentally reframe how I think about mechanisms and mechanical systems. I’m looking forward to extending this line of thinking in the future.
  • Former Post-Doc, now Research Scientist collaborator Dr. Xiaolong Liu continued pushing forward some of our joint NIH work on patient-specific tissue engineered vascular grafts, such as through his TBME paper this year that showed a full pipeline for imaging, automated design, 3D electro-spinning, and implantation of TEVGs, in partnership with our other NIH PIs, medicial roboticist Dr. Axel Krieger, cardiologist Dr. Laura Olivieri, and and surgeon Dr. Naru Hibino. We have more papers in the pipeline for this coming year, including Xiaolong’s recent medRxiv paper on robust design of TEVGs.
  • Former Post-Doc Dr. Fatemeh Ghoreishi published two papers this year as part of her MRC Fellowship both focused on Design and Optimization of Soft Robotic Catheters, with with Co-Advisors Dr. Ryan Sochol and Dr. Axel Krieger, one in the IEEE Transactions on Medical Robotics and Bionics and another in Frontiers in Robotics and AI. These papers captured her approach to the design and control of multi-actuator soft robotic catheters under various manueverability constraints for applications in cerebrovascular surgery. It was great fun working with her and seeing how her work has laid the foundation for some new work to come in future years between Ryan, Axel, and myself that I can hopefully talk about next year.
  • Some of the initial work from our ongoing ARPA-E Inverse Design of Heat Transfer Surfaces (INVERT) project has started to come out now, including Qiuyi Chen’s first paper on “Inverse Design of Two-Dimensional Airfoils Using Conditional Generative Models and Surrogate Log-Likelihoods” as part of JMD’s Special Issue on AI in Design. I found this particular paper fascinating due to novel connections between GANs and Optimal Transport measures, such as Sinkhorn Divergences. I had always been interested in OT topics but this was the first paper where I really had a chance to dig further into them on an application, and I was impressed by how well they work! In addition to this, Post-Doc Dr. Jun Wang pulled off an impressive feat to construct the 2D airfoil inverse design dataset for that paper, which you can find on our GitHub. Keep a look out for more of our Inverse Design work to come in the next year, including some work on Meta-Material ID and a some new applications in the pipeline.
  • Ph.D. student Sangeeth Balakrishnan continued some of our molecular design work via his Molecular Informatics paper published this year that studied the use of joint property embeddings to improve low-dimensional molecular graph embeddings. This is part of our ongoing ONR work with Co-PI Dr. Peter Chung, which also included closing out an article from Former Post-Doc (now Assistant Professor) Dr. Zois Boukouvalas on Independent Vector Analysis of molecular data.
  • Closing up some of the work from our Creativity Ratings NSF grant with the talented Dr. Scarlett Miller and Dr. Sam Hunter (in partnership with former Ph.D. student Dr. Faez Ahmed and Scarlett’s students), our two papers that I mentioned in last year’s update finally went to print in March and June 2021 issues of JMD, respectively: MD-20-1412 and MD-20-1087.
  • Lastly, on the education side, UMD’s use of the PrairieLearn tool has been growing. We deployed it last year via a Provost Teaching Innovation Grant, and since then the site has gone live and has been used extensively in Dynamics (ENES221), Statistics (ENME392), and my ML course (ENME440/743) in 2021. Now it is growing to include more ENES courses in 2022, and is serving as the foundation of a pilot Computer-Based Testing Facility (CBTF) for the college. What most excites me about the deployment is that it has the potential for us to leverage some of the ML tools we use in the lab to really target and improve learning outcomes for students, both at the formative and summative assessment stage. There are many neat education and research angles that I’m looking forward to exploring in the years to come.

While this was just a brief snapshot of only some of your work this year, I wanted to let you know how much I appreciate the work your all do and how it ties together the many threads of research we do in the lab. Certainly being able to interact more in the latter half of this year has been a wonderful change of pace that has helped spur complementary benefits among our different research projects.

I also wanted to remind everyone that all of the above would not have been possible were it not for the support of our various research sponsors that, at various parts of 2021, included eleven(!) projects across the NIH, NSF, ARPA-E, ONR, and the State of Maryland (via the TedCo program). So I do thank you all for your many efforts – big and small – to help deliver on the promises we make to these sponsors and the missions they support.

The Field

Oh boy, was this a banner year in terms of opportunities at the intersection of Machine Learning and Mechanical Design/Optimization, and I’m probably not even aware of all the things going on at the moment. First, the Journal of Mechanical Design held a Special Issue on “Artificial Intelligence and Engineering Design” of which Qiuyi’s article was a part. In addition, I was part of the organizing team for an upcoming workshop on “AI in Design and Manufacturing” (ADAM) that will take place at the AAAI-22 conference, with a bunch of really neat papers from a broad intersection of groups and fields, including a short paper from new Ph.D. student Arthur Drake entitled “An Epsilon-Frontier for Faster Optimization in Nonlinear Manifold Learning”. I think it is a great and healthy sign that there is a growing interest in ML and Engineering Design, not just from “traditional” Engineering Design venues, but also from Computer Science venues such as top-tier conferences. This can help grow the community and impact beyond just my “home” conferences.

Moreover, the ARPA-E DIFFERENTIATE program is in full swing now with something on the order of 23 project teams all investigating ML applications to design. I’m excited to attend the ARPA-E Innovation Summit in this coming March to see some of the other team members there and learn about what they have learned and done. Due to ARPA-E’s investment there has been increased interest in ML and Engineering Design from the National Lab community, which is also an important scientific growth pathway for the future of the field and for individual students interested in that career path. Lastly, this year I honestly lost count of the number of “ML + Design or Manufacturing” faculty openings that I saw available this year, making this an excellent time to be on the lookout for those types of positions if you want that career path. As always, if you hear of any up-and-coming researchers that you think I should add to my reading lists, let me know since I’d like to support and keep abreast of the best work in this growing area.

Alumni News

Occasionally, I’ll get some updates from past lab alumni that I like to share with the lab family each year. Specifically, Faez and his students lab at MIT have been publishing some great work in the past year, so I encourage you to check out some of their papers. Wei moved from Siemens to Northwestern University for a Post-Doc as he shifts from his industrial research career to pursuing an academic one; this has afforded him the opportunity to put out a bunch of really neat papers this year with both old and new collaborators that you should definitely check out. Zois is doing well at AU with the launch of the new Ph.D. program there and has continued his collaborations with Peter and I, so it is great to see him regularly. Xiaolong successfully secured his first grant as a PI via the Maryland TedCo program to help commercialize some of our NIH-support research. Dan Elton has moved from the NIH to Massachusetts General Hospital’s and Brighman Womens Hospital’s Center for Clinical Data Science. David Anderson’s startup Engora Inc. continues to grow and you can read about it in a recent article from the Massachusetts Tech. Leadership Council. Thanks to all the alumni who occasionally report back to me on their changes over the year so that I can share many of your updates with those who follow in your footsteps.

Lastly, I just wanted to end this letter with a note of my deepest thanks to both current and former lab members for all of your hard work that led to my being awarded tenure at UMD earlier this summer. My entire career has been built on your tireless efforts, so that milestone is much more a testament to your own abilities than it is to my own. I feel truly blessed to work with such wonderful people that I get the privilege of calling my “lab family.”

Thanks again again for what you have been able to accomplish this year, and I look forward to what 2022 will bring.