Daniel Rakita: Teaching Robot Manipulators to Effectively Move in the World Around Us

The School of Engineering & Applied Science is proud to welcome its newest faculty members for the 2022-23 academic year. The large influx of faculty members – 13 so far, with more to be announced soon – marks the rapid growth of the School and investment in the research areas illustrated in the SEAS Strategic Vision.

The latest faculty arrivals are valuable additions to the chemical and environmental, computer science, and electrical engineering departments. Their expertise includes sustainability, artificial intelligence, robotics, quantum computing, cybersecurity, and optoelectronic materials.

Upon their arrival, we asked these new faculty members questions about their work, their motivations, potential collaborations, and much more:

Daniel Rakita, Computer Science


Appleton, Wisconsin

Prior academic history:

  • 2012: Bachelor of Music in Performance – Indiana University
  • 2017: Master's of Computer Science – University of Wisconsin-Madison
  • 2022: Ph.D. in Computer Science – University of Wisconsin-Madison

How would you summarize your research?

At a high level, I develop algorithms that allow robot manipulators to effectively move in the world around us. Unlike people, robots have no common sense or intuition for their own motions. If care is not taken, a robot would naively drive its elbow into a table or over-rotate a cup and spill water on the floor. Thus, robots must be given a notion of what "good" motion means for a certain task, as well as algorithms for how to compute a motion that achieves these desired properties. For example, I have presented algorithms that efficiently generate robot motions that avoid self-collisions, avoid collisions with obstacles in a dynamic and cluttered environment, are viewed as rational and understandable by a human collaborator, and ensure that the robot's hands are in the correct positions and orientations at the right times in order to successfully execute a task. These approaches use interdisciplinary techniques across robotics and computer science, such as optimization, planning, and machine learning.

Using these motion algorithms as core components, I develop and evaluate applications on real robot systems. These systems are designed to enable many people, even those who are not experts in robotics, to intuitively control or work alongside robot manipulation platforms to perform critical tasks deemed unsuitable, undesirable, understaffed, or unsafe for people, such as full-time homecare, home assistance, telenursing, robot surgery, disaster relief, large-scale manufacturing, nuclear materials handling, and space robotics.

What inspired you to choose this field of study?

I have actually been interested in the overall concept of computationalized motion for most of my life. I remember watching Toy Story when I was five years old and being absolutely astonished by how much emotion the animators could reflect on screen via the motion of relatively simple face and body rigs. From that point on, I absorbed any information I could about computer animation. For instance, even when I went to college for music performance, I dedicated a lot of time to self-learning computer graphics. I started by learning the artistic practitioner side of using computer animation software, but over time, I gained more and more interest in its underlying computer science and math principles. My growing interest in this area led me to pursue a PhD in computer science. I started working with Mike Gleicher, a world-renowned professor in computer graphics at UW-Madison. Right around this time, Mike had been starting to apply his expertise in motion algorithms to human-centered applications in robotics. He taught me that the animation techniques I had been thinking about also apply really well to robotics; both systems are just hierarchical chains of links and joints, after all. At this time, I met one of Mike's main collaborators, Bilge Mutlu, an expert in Human-Robot Interaction, who shortly thereafter became my second advisor and forever solidified my interest in creating algorithms and applications in robotics that can better peoples' lives. And, as they say, the rest is history.

Where do you see the field 10 years from now?

I believe the evolution in robotics, as a whole, will be driven by two primary factors: (1) Tremendous advancements in machine learning and computer vision will allow robots to better interact with people, recognize their environments, and seamlessly navigate complex situations; and (2) Robot manipulation platforms will become more affordable and, in turn, become increasingly more present in peoples' homes and workplaces. These new robots in the world will refocus efforts on developing robotics technologies that are certifiably safe and usable by a broad population of people of various skills and levels of expertise. Also, roboticists will have access to a proliferation of data collected from the new fleet of robots in the real-world, nicely contributing to the machine learning and computer vision advancements mentioned in point 1.

What brought you to Yale?

When I think of Yale, the first thing that comes to mind is the people. The computer science department at Yale, as well as its related areas, is associated with some of the most knowledgeable and frankly the kindest people I have ever encountered in my young career. Throughout the interview process and post-interview visits, I consistently recognized how fortunate I would be to work as a colleague alongside the people here.

I was also so excited to hear about all the opportunities for collaborative research at Yale. Many people I spoke with indicated that there are no boundaries between departments at Yale, and interdisciplinary research is not just encouraged here, but is a "way of life". This open-ended feel to research perfectly fits my learning and research style. My mind is continuously buzzing with ideas regarding possible collaborations in which I could lead or take part. I am thrilled to have the opportunity to work with experts in my department and across the University to solve open, challenging problems in robotics and beyond.

Lastly, I am so excited about the recent enthusiastic investment in applied sciences at Yale. I resonated very strongly with the long-term vision in these areas specified by the Deans and administrators I spoke with here, and I am honored that the courses I develop and research I conduct can play some part of this expansion and vision going forward.

What areas outside of Computer Science do you seek to create impactful research collaborations or partnerships?

First, I am looking forward to fostering conversations with the Yale Medical School to see how our robotics solutions can possibly help in hospital or homecare settings. For instance, robot manipulation platforms may be able to assist with feeding, fixing of bedding, disposing of items, doing intake diagnostics tests, etc. Such assistance may be able to alleviate some of strain on overworked nursing staffs, allowing nurses to focus on tasks that more require their training and expertise.

Also, recent work has shown the potential of people being able to control robotic limbs with their thoughts through what is called a Brain-Machine Interface (BMI). I would be thrilled to collaborate with neuroscientists in the Wu Tsai Institute at Yale to see if we can make progress together in this exciting space. Generalizable solutions to this problem could substantially improve the lives of individuals with limb-loss or paralysis.

Further, there are many mathematical connections between robot motion planning and some problems related to cancer treatment research, e.g., protein folding. I am highly motivated to connect with faculty at Yale who specialize in biomedicine to see if any of the general purpose, highly efficient path planning algorithms we will continue to develop for robot motion planning in my lab can also be applied to medical fields.

Are there any courses that you look forward to teaching/creating?

I am excited to develop and teach a course on Applied Optimization and Planning in Robotics. This course will introduce students to optimization and planning techniques, such as linear programming, non-linear optimization, 2D planning on a grid, high-dimensional sampling-based planning, etc. This class will end with a visual, creative, and interactive project where students implement a simulated robot to perform tasks. I believe students will be able to transfer the knowledge gained in this class to many areas of CS and beyond.

What are your interests outside of the lab?

I really enjoy playing and watching soccer, doing computer animation, watching movies, playing board games (especially chess), playing table tennis, and reading. I am also very passionate about music – I have been surrounded by music my whole life and studied music performance in college for my undergraduate degree. A lot of my time is spent singing, playing guitar, playing violin, and playing percussion. I would be thrilled to possibly get involved with the fantastic Yale School of Music in some way, and I will certainly attend many shows, musicals, and concerts in the area with my wife, family, and friends.

What is the best New Haven Pizza?

Right now, it's Frank Pepe's, but I admit that I have not tried enough places to make a fully informed decision yet. I am really looking forward to trying it all!

Back to the 2022-2023 New SEAS Faculty Profiles