Robotic Swarms in Action

For her independent project, Sayantani Bhattacharya (MSR '25) designed a fleet of autonomous quadrupeds to navigate hazardous terrain in disaster areas.

autonomous quadrupeds
autonomous quadrupeds

A disaster zone is no place for hesitation. When every second counts, navigating treacherous landscapes such as dense forests or deep, dark mines can mean the difference between life and death.

That urgency inspired Sayantani Bhattacharya (MSR '25).

For her winter quarter independent project in Northwestern Engineering's Master of Science in Robotics (MSR) program, Bhattacharya developed and deployed a collaborative quadruped fleet using Unitree GO1 and GO2 robots – each equipped with the ability to autonomously explore unknown environments using simultaneous localization and mapping (SLAM).

“In times of disaster, reaching survivors in hazardous terrains poses significant challenges,” Bhattacharya said. “Quadrupedal robots can autonomously explore unmapped areas, merge their findings, and provide assistance and hope.”

She faced a daunting challenge in creating a robotic fleet useful in such situations. While GO1 robots required system integration across multiple peripherals and legacy code, GO2 – a newer model – demanded custom-built navigation and mapping implementations.

Beyond that, finding compatible robot operating system (ROS) packages, configuring middleware, and ensuring seamless data synchronization required meticulous troubleshooting.

Bhattacharya said she owes much of her success on the project to the foundation provided by MSR’s curriculum, particularly the Embedded Systems in Robotics course taught by MSR co-director Matthew Elwin.

“All the concepts taught in the class were crucial to this project, and I am sure they will be useful in a lot of my future work as well,” Bhattacharya said. “It was a very hands-on course, which gave me the confidence and technical know-how to take on this project.”

Sayantani's Independent project
Sayantani's Independent project

Beyond the technical aspects, Bhattacharya structured her project with a modular design, ensuring that individual components could be easily integrated into other robotics applications.

That approach reflects her broader vision of scalable multi-agent systems – an area she hopes to specialize in professionally.

Her project also involved developing a decentralized communication framework, allowing multiple quadrupeds to share and merge individual maps without relying on a central system. This enhances the robustness of search operations, ensuring that the failure of a single unit does not compromise the entire mission.

The project gave her deeper respect for the high-quality work that already exists within the robotics community. Open-source ROS packages — as well as contributions from past MSR cohorts — were indispensable resources.

“I got to experience the vast body of reproducible high-quality work that is out there for everyone to use,” Bhattacharya said. “I developed a newfound appreciation for ROS and respect for the contributors of ROS packages, those who diligently answered questions in discussion forums, and the senior MSR cohorts.”

Bhattacharya hopes to apply lessons learned from the project to the rest of her MSR studies and her future career. Her goal is to work as a robotics engineer in a fast-paced, product-focused company with a long-term vision of developing scalable, multi-agent systems.

MSR, she said, is preparing her well. The program’s emphasis on project-based learning helps simulate the complexities of real-world challenges.

“These are not just academic exercises but they mirror the complexity and constraints of real-world problems,” Bhattacharya said. “The cohort model adds an extra layer of depth, as we grow alongside peers from diverse backgrounds, constantly learning from each other’s perspectives and experiences.”

McCormick News Article