How Triton Droids Started
- sytran1
- Jan 24
- 5 min read
The idea for Triton Droids, UCSD's first and only humanoid robotics club, was born out of a passion for innovation and a desire to disrupt existing paradigms in technology and global equity. Founded by Abhinav, a third-year mechanical engineering student along with Deepta and Uday, Triton Droids began with a vision to build low-cost, highly adaptable humanoid robots designed to bridge economic disparities worldwide. Abhinav and his founding team, a dynamic mix of students from mechanical engineering, bioengineering, computer science, and business, set out to address a “Big Hairy Problem”: the inequities in production efficiencies created by the uneven distribution of capital, technology, and labor resources. According to ARK Invest, the expected value to the global GDP that will be added by humanoid robots will be $24 trillion. Given the size and scope of this future industry, the team wanted to create a club that would give students an opportunity to work on this type of problem at a university level, as there was no other club like this. They envisioned a future where humanoid robots could empower local industries in underserved regions to compete globally, reducing dependence on cheap labor and creating pathways to disrupt poverty cycles.
In the initial stages, the founders’ commitment to building an inclusive team and securing resources laid a solid foundation for Triton Droids. One of the key early successes was setting up a stall for tabling on Library Walk, which led to over 150 applications from very talented people. This recruitment drive brought in some of the best minds on campus, adding tremendous value to the project. A team of over 40 students was formed, bringing together diverse skills and expertise. Furthermore, the team secured 24/7 access to the Envision Makerspace, allowing for continuous development and prototyping. Triton Droids also formed an advisory board consisting of multiple professors from different fields, providing valuable insights and guidance. In addition, the team reached out to companies, several of which showed interest in supporting the project. The team sought to develop a robotics culture grounded in collaboration, creativity, and resilience. Realizing the potential of humanoid robots to impact multiple sectors, from manufacturing to healthcare, the team focused on creating robots that could operate effectively in any environment, regardless of access to advanced resources. The team decided to start by working on the lower body of the robot, focusing on developing robust locomotion capabilities as a foundational element for future progress.
Through recruitment drives, campus outreach, and partnerships with faculty, Triton Droids began to take shape. The team has planned to host General Body Meetings (GBMs) next quarter to keep students engaged and informed. Additionally, they are developing a timeline to complete the humanoid robot by early 2026. Triton Droids is giving students an opportunity to work on ambitious robotics challenges at a university level, inspiring the next generation of engineers and innovators.
Workday Recap
Our recent workdays on November 1 and 9 were full of energy and focused technical progress! On Workday 3, we introduced project managers, outlined project parameters, and split into sub-teams to tackle specific deliverables. The Control and Software teams coordinated on essential control algorithm requirements to ensure they’re on the same page as development begins. The Software team then divided into simulation and data communication groups to take on distinct tasks, while the Mechanical team got hands-on with CAD designs in Onshape, using real-world models for inspiration. These sessions brought our project direction into sharper focus, setting a solid foundation for the work ahead.
On Workday 4, each team dove into specialized tasks with renewed clarity. The Mechanical team worked on designs for hip, joint, and knee mechanisms, further developing their CAD models to meet our project’s specifications. Controls took things up a notch by simulating PID control strategies in Simulink, and the Software team prepped for ROS2 installation by setting up Docker environments for consistency. Meanwhile, the Simulation team split to evaluate MuJoCo and Matlab, working to understand the strengths and weaknesses of each platform for our needs. We’re making tangible progress with every session, building toward an exciting final product, and the teamwork is stronger than ever!
Faculty associated with Us
Our faculty advisor, Dr. Michael Yip, is a leading expert in robotics, offering extensive support and guidance to our project. Dr. Yip’s focus on advanced control systems and robotic manipulation has been instrumental in helping us navigate complex challenges as we design and build robot legs. His insights will sharpen our approach to developing robust, responsive, and adaptable leg mechanisms capable of dynamic and stable movement.
Complementing Dr. Yip’s expertise is our growing advisory board, which includes Dr. Nicholas Gravish and Dr. Xiaolong Wang. Dr. Gravish’s extensive knowledge in biomechanics and bio-inspired robotics provides us with a unique perspective on how nature’s principles can enhance mechanical efficiency, adaptability, and resilience in robotic designs. Through regular discussions with Dr. Gravish, we want to explore ways to mimic and adapt biological strategies in our robotic leg designs, such as optimizing energy efficiency during locomotion and enhancing stability over uneven terrain.
Dr. Xiaolong Wang brings a strong focus on reinforcement learning and robotic perception, which are crucial for enabling our robot legs to interact intelligently with complex environments. His expertise helps bridge the gap between mechanical design and intelligent control, encouraging us to think beyond rigid programming and consider adaptive, learning-based strategies that enable our robots to improve their performance over time through interaction and experience. Our discussions with Dr. Wang will delve into strategies for integrating real-time feedback and data-driven decision-making capabilities into our designs, enabling the legs to adjust to diverse scenarios with precision and efficiency.
Together, Dr. Yip, Dr. Gravish, and Dr. Wang form a well-rounded team of advisors who contribute complementary strengths and deep insights into the various technical aspects of building robot legs. Their combined guidance ensures that our project is grounded in both cutting-edge research and practical design principles. This collaboration not only accelerates our progress but also inspires innovative approaches that push the boundaries of what robotic systems can achieve. As we continue to iterate and refine our designs, their mentorship fosters a learning environment where we are encouraged to tackle ambitious goals with confidence, knowing we have the expertise and support to achieve excellence.
Gabriel’s Experience at the SHPE Convention at Anaheim
Hello everyone, I am Gabriel Bortoni, a part of the controls and sensors team at Triton Droids. I was able to attend the 2024 SHPE National Convention, a four day convention that seeks to empower hispanics in STEM, while helping them jumpstart their careers. In the convention, I was able to meet a lot of people interested in robotics, ranging from undergraduate students to PhD students to full-time employees at different companies.
First, I met with Jairo Maldonado-Contreras , a Robotics PhD candidate at Georgia Tech. He granted the Triton Droids team valuable advice on how to approach the initial stages of development for our humanoid robot. With his robotics expertise, he recommended several simulation tools and research papers that we could look into for testing and brainstorming potential control structures.
Second, I spoke with employees at Ansys, a simulation software design company, who kindly offered their services to aid our search for simulation software that we could use to test and develop our humanoid robot in a cost-efficient manner.
Last but certainly not least, I visited the Dominion Energy booth at the convention’s career fair and interacted with their one-of-a-kind employee, Sparky. Sparky is a Boston Dynamics robot dog, also known as Spot, that is controlled by Joshua Guzman Bell, a Nuclear Technology & Innovation Consultant at Dominion Energy. This robot dog helps conduct autonomous inspections at a commercial US nuclear power station owned by Dominion Energy. Sparky’s job could be achieved by a human; however, these inspections often lead to exposure to radiation, which threatens human safety. Therefore, Sparky provides a safer, more efficient alternative that gets Dominion Energy the data and information that they want from its locations without putting anyone in harm's way.
Comments