NASA JPL Internship
I got the opportunity to intern in the Jet Propulsion Laboratory's Robotics section. I worked on a robotic inspection system, which included a quadruped and a tether system. I had three objectives throughout my time there: Create a mass optimization proposal based on the roller arm, evaluate the inspection capabilities, and operate the test bed and provide hardware support.
Robotic Platform
The quadruped is a unitree A1 customized with a roller arm which helps the robot move through distillation tower manways. The arm is actuated by a timer pulley, with 3D printed base and aluminum motor housing. The arm has actuated urethane rubber rollers to help with the transition between tower layers.
Mass Optimization
I was able to find the same performance with a lighter system by changing the gearbox of the wheel actuators. This made the actuator less robust, I added a bearing to the motor housing share the radial load. This motor housing was also optimized by adding a fillet instead of the cyan extrusions.
Further mass optimization was achieved by redesigning the roller wheels. The aluminum hub was optimized using SolidWorks's topology optimization , and the design's machinability was verified by Fictiv, a CNC manufacturing company. The modifications are projected to result in a weight reduction of 1.3 kg, equating to a 27% decrease in the overall mass of the robot.
Inspection Capablities
The work done before I came was mostly based on the localization and locomotion inside the column. One of my initial tasks was to generate a document outlining the extent of the column that our existing system can inspect and proposing solutions for tasks beyond its inspection capabilities.
​
I did this by creating a 2D rendering of the column and robot to check the geometric constraints. I then upgraded this by using a 3D model of the robot and a model representing the smallest column it will have to inspect.
Testbed Operator and Hardware Support
I tested the robot in the mock column shown on the left. I ran tests for load capacity, temperature, and communications.
Throughout the testing process, I documented my observations, identifying potential issues and proposing sources for them.
I created a voltage enclosure for the test bed and tested it's effectiveness while running the robot for extended periods of time. I chose to use PLA because the melting point of it was significantly lower than the max temperature of the voltage converter. I tested my design by using a thermocouple to confirm cooling.
JPL Experience
Working at JPL was amazing and I absolutely loved how they treated their interns and placed a good amount of responsibility on them. They had professional workshops, networking events, and had new speakers every week which were experts on their given field.