Back in Spring 2019, a couple of my engineering friends and I started a student research team called Tufts Exosuit, which was created with a goal to attempt to design a soft exoskeleton or exosuit apparatus to help correct sedentary posture and reduce lower back pain. Excited as we were, we started looking for possible directions to get the project started, mainly on the Soft Robotics Tookit website.
As we continue our research, we quickly realized the tremendous amount of work needed to bring our original ideas into reality in some form - in the fields of both soft robotics and biomechanics. Therefore, we decided to divide the team into two subgroups, with me and Maya Kurzman, E'21, leading a team to explore in the field of soft robotics, and William Liu, 'E21, leading the biomechanics team to conduct usability testing on sitting postures. We hope that the two sub-teams can work more closely in the future to come up with an amaing Exosuit product.
Currently, my work within the research group is primarily focusing on design for manufacturing simple silicone actuators powered by compressed air, as well as various methods to control them.
My undergraduate independent research course: Developing a low-cost, modular pneumatic system for controlling soft robot actuators.
From replicating an open-source soft robotics control board to making my own version with improvements
This modular station allows you to play with up to 10 development boards at the same time with only one input device.
As a passionate maker, staying at home due to a pandemic was frustrating at first. Luckily, I was finally able to force myself to improve on what used to be my biggest weakness - coding. In the summer of 2020, I was very lucky to work remotely as an intern (again) at Tufts Center for Engineering Education and Outreach (CEEO). Along with a small group of talented students, we worked closely with LEGO Education to explore possibilities to use LEGO Education SPIKE Prime Set to teach middle and high school students artificial intelligence - mainly machine learning.
The job seemed intimidating to me at first, considering that I mainly study mechanical engineering and know little to nothing about AI. With help from my team leads, I went through a couple of AI crash courses and was able to catch up. With my expertise in building LEGO robots, I was able to come up with various demo activities that uses the LEGO SPIKE platform to teach machine learning concepts: K Nearest Neighbors, K Means Clustering, Support Vector Classification, etc. I was also able to build and test a LabVIEW user interface that integrates LEGO SPIKE Prime smoothly with scikit-learn, an advanced machine library available in python environments.
As a continuation of my summer work, I am currently working with a PhD student at the center to write a paper for the AAAI Conference on using the LEGO Robotics to introduce reinforcement learning concepts to middle school students.
