Aditya Sripada (left) and Abhishek Warrier (right)
Curiosity allows humans to pull ideas from fiction and give life to them in the real world. For Carnegie Mellon University alumni Aditya Sripada and Abhishek Warrier, a question inspired by a fictional robot in the film “Interstellar” served as the impetus for building, testing and ultimately reenvisioning how robots can move.
The result was TARS3D, a minimalist, shape-shifting robot that combines the efficiency of wheels with the adaptability of legs. The project went on to earn recognition at the 2025 IEEE-RAS International Conference on Humanoid Robotics, where it was selected as a finalist for the Mike Stilman Award. The award, which honors the legacy of Robotics Institute alumnus Mike Stilman, annually recognizes conference papers that demonstrate notable originality and innovation in humanoid robotics.
Sripada, who earned his master’s in robotics in 2021, and Warrier, who earned his master’s in robotics systems development in 2025, both now work at Nimble Robotics in California. Below, they discuss how a science fiction idea evolved into a real-world robot, what motivated them to pursue the project outside a formal lab and how their time at Carnegie Mellon shaped their approach.
How did starting with a pop culture inspiration affect your creative process?
Aditya: “Interstellar” is my favorite movie of all time. In the film, the TARS robot is a unique, all-terrain robot that can both walk and roll. But because the movie used a mounted prop for those scenes, there was no real precedent for how such a robot would move in real life. I always wondered if it truly was something we could accomplish in the real world. Approaching the project from that “Is this even possible?” angle meant we weren’t constrained by existing robotics conventions. Instead of trying to improve on a standard legged robot, we were trying to bring a fictional mechanism to life. That gave us permission to think more experimentally.
You two worked on this project during your free time without any funding or formal affiliations. How did you stay motivated?
Aditya: For us, motivation came from pure curiosity and enjoyment. This started as a passion project –– just wanting to build robots again and finally explore a question that had been lingering for years. I had a bunch of actuators lying around from another project, and had just bought a new 3D printer, so we accomplished rapid prototyping and testing rather easily. Working without a funded lab meant we had to engineer creative solutions to challenges that normally require specialized, expensive equipment — which honestly made it more rewarding. It felt like getting back to our roots. We are just two people building robots because we love it. That made it easy to stay motivated.
Abhishek: It also helped that we’ve known each other since undergrad in India. We were two years apart in our schooling, but we were in the same robotics lab, where we worked on locomotion projects together. Later, we both ended up doing our graduate degrees at CMU, so we maintained that connection. Even without external funding or deadlines, we pushed each other forward simply because we both cared about the problem and enjoyed the challenge.
TARS3D can both walk and roll, making it ideal for diverse terrains
TARS3D can walk and adapt its movement in ways traditional legged robots can’t. For people outside robotics, what real-world problems could a minimalist, shape-shifting robot like this help solve?
Aditya: With humanoid robots, we often see that walking tasks are ideal for navigating uneven terrain or stepping over obstacles, but it takes a lot of energy to just stand up and maintain balance on their legs. On the other hand, wheels are super efficient for getting from point A to point B, but they can’t handle discrete terrain the way legs can. So a robot that could both walk and roll would be incredibly beneficial. Structures like TARS3D are naturally stable. They don’t have to constantly fight gravity to stay upright, which means they can save a lot of energy while still being versatile. That combination could be valuable for applications like planetary exploration — on Mars or the moon, you need a robot that can handle unpredictable terrain without wasting energy, because every watt matters. TARS3D started as a curiosity-driven question: Can sci-fi-inspired designs actually work? By proving the concept, it opens new avenues for design exploration that the robotics community can take further.
Abhishek: The architecture and design are very simple. It only has seven degrees of freedom. It can walk and it can roll. It can maintain stability without having to rely on active motor torques at the joints. All of this would make robots like TARS3D useful in extreme environments with unpredictable terrains or obstacles.
You presented your paper at the IEEE-RAS International Conference on Humanoid Robotics, where it was a Mike Stilman award finalist. How was this experience for you?
Aditya: It was pretty surreal, honestly. It’s amazing to think that something that started purely as a passion project ended up being a finalist for an award given to the most innovative work at one of the top conferences in humanoid robotics. It felt really validating, and it was also just great to reconnect with a lot of people from CMU that I hadn’t seen in a long time.
Abhishek: I was unfortunately not able to travel to the conference in person, but it was still a really meaningful milestone. The humanoids conference is known for having a high bar and a specialized community. So getting the paper accepted and then selected as a finalist felt like a strong validation of the work we put in. It was exciting to share it with people who really understand the technical challenges behind it.
What aspects of your training or past research at CMU prepared you for this work?
Aditya: Being advised by Hartmut Geyer at CMU really helped with this experience. He’s one of those advisers who believes that how you communicate your research is just as important as the research itself. He always pushed me to explain things in a way that someone outside my specific field could still understand. Being clear, intentional and grounded in the core science really shaped how I approach problems. I learned that the best research comes from deeply understanding a problem and communicating that understanding clearly. That philosophy stuck with me and taught me how to investigate questions systematically, which directly influenced how I tackled the TARS3D project. Through that whole process, I think I really came to understand the essence of research and how to systematically approach a problem.
Abhishek: To add onto that, one thing that stands out from my time in the master’s in robotics systems development program is how much emphasis there was on technical writing. We wrote reports, documented our work and even had grammar quizzes in the second semester to sharpen how we communicated our ideas. As funny as that sounds, it really helped us to clearly explain our reasoning and our results, which helped us put together the TARS3D paper. Those skills have been one of CMU’s biggest contributions to my journey. Plus, at CMU we were trained to think about research –– the problem-solving mindset, the structure, the discipline — it all has really stayed with me. It made it easy to know how to break down a problem and decide what to do next.
Do you have any advice for other computer science students who may want to pursue a passion project but may not know where to start?
Aditya: Keep at it. I know the time and the means are hard to come by, but if you’re passionate about something, the process itself will be so rewarding that the work won’t even feel like it’s too hard or like it’s taking a lot of your time. When I first started this project, I had no expectation that it would go anywhere. It was just something I tried over a weekend, but that’s really all it takes to begin. There’s real joy in building something with your own hands, whether it’s software or hardware, and then watching it actually do something in the real world –– that feeling is one of the biggest dopamine hits you can get.
Abhishek: I think it’s never been easier to get into robotics, given how many resources and how much open information is out there now. It’s important to remember that whatever you build doesn’t need to have an immediate application or a big impact for it to be worthwhile. The real value is in taking an idea from something that lives in your head to something that exists in the physical world. Like Aditya said, the process is incredibly rewarding on its own. And remember, a lot of great work in robotics starts from simple inspiration –– sometimes even from science fiction.
For More Information: Aaron Aupperlee | 412-268-9068 | aaupperlee@cmu.edu