Swarm robotics is the study of how many simple robots can work together to create complex behaviour. It is one of the most fascinating areas in robotics, but actually getting started with it is surprisingly difficult. Most swarm robotics platforms are expensive, hard to buy, locked behind proprietary software, or built only for a specific lab.
Minnow is being built to make swarm robotics more accessible. It is an affordable, open-source kit of small ground robots designed for researchers, students, educators, and builders who want to experiment with swarms without needing a six-figure lab budget.
The goal is not to build a robot that pretends it is ready for real-world deployment. The goal is to build a proper research and teaching tool: something people can use to study decentralized coordination, neighbour communication, collective behaviour, robot-to-robot interaction, and swarm algorithms in the real world.
Inspired by projects like Crazyflie, where the value is not just the hardware, but the software, documentation, ecosystem, and community around it.
Here’s a brief overview of what I’ll cover:
The design of Minnow: a small ground robot built around a round PCB-as-chassis, ESP32-S3, differential drive, ToF distance sensors, and an IR neighbour-communication ring.
Why we are designing for affordability, with a target of under $35 per robot in small kits.
Why decentralized control matters, and why each robot should run its own logic instead of depending on a central computer.
How we use ESP-NOW for peer-to-peer communication and an overhead camera only for logging and ground-truth data, not for controlling the swarm.
The challenges of designing a robotics platform around India-first sourcing and small-scale manufacturing.
Why we think there is space for a truly open, buyable, decentralized ground swarm kit with good software and documentation.
The idea of treating Minnow as a scientific instrument, not a deployment robot, and how that changes the way we design the hardware and software.
My goal with this talk is to show that meaningful swarm robotics research does not need to be locked away inside expensive labs. With the right tools, documentation, and community, more people should be able to build, study, and experiment with swarms of their own.
