Robotic Dog
Project Introduction
The parts we were given weren’t ideal for building a robot, which made the challenge more interesting and forced me to think carefully about how to get the most out of the available components, both mechanically and electrically. I also joined the class about two and a half weeks late, and since most people had already formed pairs, I ended up working independently. That ended up being part of the challenge I leaned into, taking full ownership of the design, build, and iteration process to bring the system to a working result.
Planning
Going into the project, I knew I didn’t want to build just another basic forward-moving robot. I wanted something more interesting and dynamic, so I explored a few different concepts before settling on a final direction. One idea was a frog-like robot that would leap forward, but after some quick calculations, it became clear that the servos were too slow to make that feasible. I considered adding gearing to increase speed, but given the 3D-printed structure, I was concerned about introducing excessive stress and potential failures, as well as system rigidity issues. I also explored a more creative “self-propelling track” concept using a circular arrangement of servos, but it lacked the ability to turn and didn’t align with what I wanted out of the system.
I ultimately took inspiration from Boston Dynamics-style quadrupeds, which offered both mobility and a more engaging design challenge, along with the ability to move in multiple directions and expand into more advanced behaviors later on. From there, I moved into feasibility, sketching out the system and working through constraints like weight, limb length, and available hardware. The body size was largely dictated by my 3D printer, since I wanted it to be a single rigid piece to avoid weak points from multi-part assembly. That decision drove the overall proportions of the robot, including leg length, which had to balance speed, strength, and aesthetics. I then ran basic calculations to confirm that the servos could handle the load and achieve the target speed, while ensuring all electronics could fit within the body.
Beyond basic functionality, I also planned for durability and usability. I incorporated shock absorption in the feet to reduce stress on the 3D printed parts, minimizing the risk of layer-based fractures over time. I also considered how the robot would behave in failure cases, designing the outer structure so it could recover if flipped over. Throughout the planning process, the focus was on creating a system that was not only capable of moving forward but also mechanically sound, compact, and intentionally designed.
Design & CAD Development
The design process was a mix of sketching ideas by hand and quickly translating them into CAD. I started with rough notebook sketches, then iterated directly in CAD to refine the geometry and validate how everything would fit together.
I began with the body, since it defined the overall structure and had the most constraints. It was designed as a simple two-part system, with a bottom section that holds the heavier components, such as the battery and servos, to keep the center of gravity low and improve stability. These components were enclosed and secured with a top plate, creating a structure similar to an I-beam, which added stiffness without unnecessary weight. Lighter electronics and components that required frequent access were placed on the top, making wiring and adjustments much easier during iteration.
For the legs, I focused on creating smooth, organic geometries using lofted features to avoid sharp corners and reduce stress concentrations. This was especially important given the 3D printed construction, where layer-based weaknesses could lead to failure under repeated loading. The same approach was applied to both upper and lower leg segments to maintain strength while achieving a clean, intentional aesthetic.
The feet were printed in TPU to introduce passive shock absorption, reducing impact forces during walking and protecting the structure over time. Throughout the design, I used threaded heat inserts and screws to allow for easy disassembly and rapid iteration. The outer shell was also designed with a smooth curvature so the robot could recover if flipped over, adding a simple but effective robustness feature.
Assembly & Validation
Rather than designing the entire system and assembling it all at once, I worked iteratively, building and validating one section at a time. I started with the body, then added and refined individual components like the legs and feet, ensuring each part functioned properly before moving outward. This approach allowed me to quickly identify issues, adjust designs, and ensure that each subsystem integrated smoothly without interference.
During assembly, the primary focus was on fit, range of motion, and reliability. I tested servo movement, checked for collisions between components, and refined clearances to prevent parts from interfering with each other. Through this process of rapid iteration and validation, I arrived at a final system that performed as intended and matched the design goals.
Testing
After completing the full assembly and confirming that all components moved freely without interference, I programmed a basic gait to validate that the system could walk as intended. This was a simple, manually defined motion used to verify functionality and ensure nothing was overlooked before moving into more advanced control methods like inverse kinematics.
The robot was able to walk successfully using this basic gait, confirming that the mechanical design, clearances, and overall system integration were sound. The next step is implementing an inverse kinematics-based gait for more advanced control; due to working independently and time constraints, this was not completed during the project, but it is something I plan to continue developing.
My in-depth progress report can be found here.
Gallery
