Neato Tag

Overview

For my Introduction to Computational Robotics final project, our group wanted to experiment with both swarm robotics and multi-sensor fusion. The result was Neato Tag: a playful system where Neato robots chased and tagged each other autonomously.

Each robot combined data from a Raspberry Pi camera and its onboard LiDAR. We placed semi-circles of sticky notes on the back of the Neatos to act as tags, and bump sensors confirmed when one robot successfully tagged another. ROS2 nodes on a central computer coordinated the system, allowing multiple Neatos to play tag together.

We created a successful proof-of-concept where three Neato robots played tag autonomously, showcasing scalable swarm behaviors.

My Contributions

• Developed an algorithm to detect semi-circles of a given radius in LiDAR point clouds.
• Integrated Raspberry Pi camera feeds with OpenCV for Neato detection.
• Researched Kalman filtering for sensor fusion (though not deployed in the final version).
• Wrote the target-path algorithm combining LiDAR-based obstacle avoidance with other Neatos’ positions, feeding into a potential field / gradient descent model.

Results

The project was highly successful: we had three Neatos playing tag autonomously, with the ability to expand to more robots given hardware. The system could run continuously until battery depletion.

We also built an alternate mode: the Neato that was “it” ran autonomously, while the other Neatos were piloted remotely by humans attempting to avoid being tagged. This made for a fun demo and showcased the flexibility of the control system.

See videos here.