Original title: Synthetic Wastewater Epidemiology Data Generation using Patterns-of-Life Simulation
Hossein Amiri, Mohammad Hashemi, Akshay Deverakonda, Joon-Seok Kim +2 · August 2026
Sewage holds biological signals that can track diseases across entire communities. However, real wastewater data is hard to collect. It requires lab testing, field sampling, and expert analysis, leaving real-world datasets rare, short, and geographically fragmented. To fix this shortage, researchers created a way to generate realistic synthetic data. They built on an existing simulation framework that models human behavior and movement. They added rules for how infections spread between people and how sick individuals shed pathogens into sewage. The system tracks simulated daily life, simulates people catching illnesses, and records the pathogen signals that would end up in the wastewater system. The researchers generated full datasets for Fulton County, including infection states, social links, movement records, and pathogen loads. Because the setup works with OpenStreetMap information, it can also produce synthetic datasets for any other city with available map data.
Imagine learning to fly. You cannot practice an engine failure by actually breaking an engine, so you sit in a flight simulator that recreates the emergency safely. In this paper the same idea is applied to disease surveillance: virtual people move around a city, catch an infection, and release virus into the sewers, producing fake records that look real enough to test outbreak tools on.
Wastewater data is hard to get: it requires field sampling, lab analysis, and expert interpretation, so datasets are scarce and fragmented. This makes it hard to develop and compare outbreak-detection tools. Synthetic data with known ground truth lets researchers test those tools freely, and the framework can generate it for any city with OpenStreetMap data.
The abstract doesn't report validation of the synthetic data against real wastewater measurements, so it is unclear how faithfully the simulated signals match reality. It also doesn't compare the generated datasets to existing benchmarks or real-world outbreak data. The framework is demonstrated for Fulton County, and while it generalizes to any city with OpenStreetMap data, the abstract doesn't describe how quality varies across regions.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Synthetic Wastewater Epidemiology Data Generation using Patterns-of-Life Simulation — Hossein Amiri, Mohammad Hashemi, Akshay Deverakonda et al.
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.