Layered-Labs/nyc-clinic-ai-infrastructure

NYC Clinic AI Infrastructure

Broadband, electricity, and grid reliability data for all 311 NYC ZIP codes: can a clinic run local AI?

   

---

Overview

NYC Clinic AI Infrastructure maps three infrastructure prerequisites for on-premise AI deployment across all 311 New York City ZIP codes. Each row covers one ZIP code with fixed broadband subscription rates from the Census ACS, ISP coverage and max speeds from the FCC National Broadband Map, commercial electricity rates from the EIA Open Data API, and grid reliability metrics from EIA Form 861. The dataset enables data-driven analysis of where local, offline-capable AI is feasible for community health clinics, and where infrastructure gaps remain the binding constraint.

---

Statement of Need

Community health clinics serving underserved populations are under pressure to adopt AI tools for triage, documentation, and clinical decision support. On-premise AI keeps patient data local, reduces recurring cloud costs, and works in low-connectivity environments. But feasibility depends on infrastructure: reliable broadband, affordable electricity, and stable grid power. This dataset provides the first ZIP-level infrastructure readiness profile for all 311 NYC ZIP codes, enabling evidence-based decisions about where on-premise AI deployment is viable today and where investment in digital infrastructure is needed first.

---

Intended Use

This dataset is intended for researchers, policymakers, and health equity organizations studying the feasibility of on-premise AI deployment in community health clinics. It is designed for analysis, visualization, and policy research. It is not a real-time infrastructure monitoring tool and should not be used as a substitute for current ISP or utility data in procurement decisions.

---

Limitations

FCC broadband columns cover 197 of 311 ZIP codes; the remaining 114 have NaN broadband data due to Census ZCTA boundary mismatches. Electricity rate and reliability metrics are uniform citywide as Con Edison serves all five boroughs. FCC coverage data reflects advertised availability, not actual subscribed speeds or service quality. Census broadband figures are from the 2022 ACS 5-year estimates and may not reflect current subscription rates. Data vintages vary across sources (FCC Jun-2025, Census ACS 2022, EIA 2023-2024).

---

Dataset Structure

Splits

Split	Rows
train	311

Features

Column	Type	Description
zipcode	string	5-digit ZIP code.
borough	string	NYC borough (Manhattan, Brooklyn, Queens, Bronx, Staten Island).
census_fixed_broadband_pct	float	Percentage of households with fixed broadband (cable, fiber, or DSL).
census_no_internet_pct	float	Percentage of households with no internet access.
fcc_isp_count	int	Number of unique ISPs offering service in the ZIP.
fcc_fiber_available	bool	Whether fiber-to-premises service is available.
fcc_max_dl_mbps_practical	int	Max advertised download speed (Mbps), capped at 10 Gbps.
commercial_rate_cents_kwh	float	Commercial electricity rate in cents per kWh (EIA 2023).
monthly_cost_mac_mini_m4_usd	float	Monthly electricity cost for a 35W device running 24/7.
monthly_cost_workstation_gpu_usd	float	Monthly electricity cost for a 350W GPU workstation running 24/7.
saidi_excl_major_events_min_yr	float	Grid outage duration per year in minutes, excluding major events (EIA Form 861 2024).
reliability_tier	string	Grid reliability tier: excellent, good, fair, or poor (based on EPRI benchmarks).

---

Usage

from datasets import load_dataset

ds = load_dataset("Layered-Labs/nyc-clinic-ai-infrastructure")
print(ds)

Example

from datasets import load_dataset
import pandas as pd

ds = load_dataset("Layered-Labs/nyc-clinic-ai-infrastructure")
df = ds["train"].to_pandas()

# ZIPs with lowest broadband access
low_access = df.nsmallest(10, "census_fixed_broadband_pct")[
    ["zipcode", "borough", "census_fixed_broadband_pct"]
]
print(low_access)

# Monthly cost to run a Mac Mini M4 at each ZIP
print(df[["zipcode", "monthly_cost_mac_mini_m4_usd"]].head())

---

Citation

If you use this dataset in your research, please cite:

@dataset{layeredlabs_nyc_clinic_ai_infrastructure,
  title     = {NYC Clinic AI Infrastructure},
  author    = {Ridwan, Abdullah and Hossain, Radhyyah},
  year      = {2025},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/Layered-Labs/nyc-clinic-ai-infrastructure}
}

---

License

Released under the CC-BY-4.0 License. Maintained by Layered Labs.