Underworlds

Developing a robotic platform for real-time urban epidemiology anchored this research project at the Massachusetts Institute of Technology’s Senseable City Lab and Alm Lab. The work focused on embedded hardware and robotics, specifically building a system to autonomously collect and analyze sewage water for bacteria, viruses, and chemical compounds. This initiative aimed to map human gut health on a city-wide scale by deploying interconnected sensor networks directly into municipal wastewater infrastructure.

The core of the system is Luigi, a second-generation robot prototype designed to operate suspended beneath street-level manholes. The hardware architecture integrates a motorized pulley, a pumping mechanism, and embedded electronics powered by internal batteries. Upon deployment, the robot lowers itself to the water level and extracts approximately ten liters of sewage. This fluid is processed through an internal filtration system before being expelled, allowing the device to capture concentrated biological samples.

Operating in harsh, highly corrosive sewer environments required strict physical isolation of all electronic and mechanical components. The system had to prevent cross-contamination of the biological samples while ensuring the hardware remained easily removable for laboratory extraction. Iterating upon a previous prototype named Mario, this second-generation build prioritized a leaner, faster, and cheaper physical footprint. This simplified mechanical design allowed the robots to be manufactured in series and deployed in collaborative groups.

The resulting hardware fleet provided a reliable method for continuous wastewater monitoring, demonstrating how embedded robotics can operate within existing subterranean infrastructure. By transitioning from a single proof-of-concept to a deployable series of autonomous samplers, the project established a functional tool for public health tracking. The system validated the feasibility of distributed sewage analysis, with further project details documented at underworlds.mit.edu.