Methodology

1. About Cuba Watch Dashboard

Cuba is experiencing chronic shortages of food and hard currency, aging and poorly maintained water infrastructure, and a deteriorating power grid prone to frequent, prolonged blackouts that disrupt refrigeration, medical care, and water pumping. The island's geography and fragmented transportation network make it difficult to move aid consistently from ports and airports to the municipalities that need it, particularly in rural and eastern regions. Reliable, current information on these conditions is fragmented across open sources, inconsistent in format, and difficult to evaluate geographically or over time.

Approach

Purpose & Framework: Cuba Watch gives humanitarian organizations, U.S. Government agencies, and other stakeholders a single, geospatial and analytical understanding of the volatile conditions shaping aid delivery in Cuba. It organizes that picture using PMESII-P — political, military, economic, social, information, infrastructure, and physical environment variables — built around one goal: getting aid to the people who need it.

Methodology: Cuba Watch draws on open-source government records, international organizations, geospatial platforms, and academic/policy research to build a continuously updated, geolocated picture of each PMESII-P category across Cuba's municipalities.

Objectives

• Cuban population in the greatest need of food, clean water, sanitation, and medical support identified.
• Accurate picture of critical infrastructure, distribution routes, and security openly shared.

Overarching Research Questions

• Where are the populations in greatest need of food, clean water, sanitation, and medical support in Cuba?
• What are the locations and status of the critical infrastructure needed to support delivery of aid to those populations?
• What is the security environment in those areas?

The platform serves humanitarian organizations and NGOs (verified site-level data for routing and prioritization), military and defense planners (pre-crisis contingency planning and logistics), and intelligence and policy analysts (a structured, sourced, confidence-rated infrastructure database tied to transition-scenario planning).

2. Data Sources and Collection

Cuba Watch draws on open-source government records, international organizations, geospatial platforms, and academic and policy research to build a continuously updated, geolocated picture of each domain across Cuba's 168 municipalities. Site-level collection followed three tiers, identifying primary and alternative sites per municipality by capacity, functionality, and suitability for aid distribution, with each postal address then verified. Most of the data was collected by the FIU analysts, searching various, reliable websites (listed per variable) and geolocating each site for additional verification.

3. Unit of Analysis

The dashboard’s unit of measure is at the municipality level. Cuba comprises 168 municipalities across 15 provinces and the special municipality of Isla de la Juventud. Point features carry their own coordinates and are attributed to the containing municipality.

4. PMESII-P Framework and Dashboard Layers

Each numbered category and subcategory below is a dashboard layer or toggle that can be turned on or off within the legend. Display, features, and variable symbols depend on the platform the dashboard is integrated into. Each variable has its own coding rule, then its layers with symbol, OSINT availability where assessed, analytic purpose, data source, and current data status.

All domains draw on open-source and independent media reporting, human rights and event monitoring, satellite imagery, verbal communication with in-country sources, and academic and policy research.

5. Site-Level Assessment Framework

This establishes which facilities within a municipality can receive and distribute aid, and how they rank. Sites are mapped not as isolated points but as components of distribution networks, incorporating capacity, conditions, transit corridors, and entry points.

5.1 Infrastructure types

Three categories are tracked, selected for operational relevance to humanitarian distribution and data availability for Cuba. Each is a node type in the aid delivery system; together they cover the supply chain from entry to delivery. All three are coded Social as community-facing distribution sites, with physical plant attributes cross-referenced to Infrastructure. This part of the data was gathered and reviewed alongside the OAA team, following a three-tier process.

• Tier 1 — Directory identification: Site listings from the Cuban White and Yellow Pages (Páginas Amarillas), the main publications of ETECSA, Cuba's state telecommunications monopoly, to obtain postal addresses and phone numbers. Facilities were independently verified using Google Earth satellite imagery, maps, and verbal communications.
• Tier 2 — Expert curation: OAA experts curated hospitals, schools, and religious sites suitable for distribution, based on conditions and capacity including transportation access, cold storage, and functioning generators, producing the Humanitarian Utility Score (Section 6.5). Their assessment draws on more than 30 years of field operations, review of local media, and direct knowledge of aid delivery across Latin America and the Caribbean.
• Tier 3 — Geospatial authentication: FIU analysts matched addresses to Google Maps listings to authenticate location and assign coordinates, cross-referencing name, municipality, and province against Páginas Amarillas records. Each mapped site links to both records so distributors have directions and contacts. Sites below the minimum confidence threshold were excluded and are not displayed. Population per municipality was gathered from the 2022 Cuban census.

Table 1. Infrastructure Types

5.2 Hospital refrigeration capacity

All Cuban hospitals and large polyclinical facilities have refrigeration. All were built, rebuilt, and are operated by the Ministry of Public Health (MINSAP), following the constructive model called Girón-I, Girón-II, or Girón modified. National-category hospitals have refrigeration (“cuarto frío”) of 150 cu m (1,633.5 cu ft); provincial level 100 cu m (1,089 cu ft); municipal-level hospitals and polyclinics 50 cu m (544.5 cu ft). These sizes are standard in that design. Smaller units preserve blood plasma, vaccines, and other pharmaceuticals; they are not commonly used for food or perishables, though well-packaged foods can be preserved. Larger ones have specialized areas for medicines, chemicals, and food. Confirmed by verbal communication with a source who manages the MINSAP database, whose identity cannot be shared.

Note: Cold chain is grid-dependent — linked to electrical dependency and outage measures in Infrastructure (Section 5.6).

5.3 Hospital electricity generator capacity

All hospitals and large polyclinics have generators, an overwhelming majority replaced after the COVID pandemic. Most are made in Spain, some donated by Spain; all diesel, typically Dagartech or Himoinsa. Individual capacity is unknown, suspected between 10 and 50 KVA, but sources report good condition given they are approximately three to four years old. Older but operational generators were donated to Poder Popular - Comunales offices. The biggest electric concern is lack of fuel to operate them under frequent blackouts; sources report certain areas have greater electricity problems than others, notably the eastern provinces.

Note: Backup generation is meaningful only against fuel availability (Economic, Section 5.3) and outage duration (Infrastructure, Section 5.6). A generator record read alone overstates site resilience.

5.4 Transportation access categories

Assessment was based primarily on verbal communication and satellite imagery, supplemented by local newspapers and municipal government information. Cuba's road and railway networks are generally in fair or poor condition. Major roads were originally designed for loads up to 50 metric tons including bridges; secondary roads generally support 30 metric tons, though bridge conditions may create bottlenecks.

Table 2. Transportation Access Criteria

Note: The site-level expression of Physical Environment (Section 5.7). Road status changes propagate to a site's category and through it to its Humanitarian Utility Score.

5.5 Humanitarian Utility Scoring

The Humanitarian Utility Score (HUS) is a composite indicator developed by the OAA expert team to support operational prioritization. It synthesizes four site-level dimensions — verified capacity, structural condition, accessibility, and network position within the distribution chain — into a single normalized score enabling planners to rank facilities by readiness and utility. A high HUS reflects a facility that is functional, reachable, and well-positioned; a low score may indicate constraints in condition, access, or isolation from distribution routes. Low-scoring examples include the La Felicidad and El Palenque communities in Yateras municipality, which depend on a small road and several potentially precarious bridges and culverts. The HUS is updated each verification cycle and should be read alongside each site's confidence rating and staleness flag. It informs operational judgment; it does not replace it.

Table 3. Humanitarian Utility Score Criteria

Cross-reference: HUS measures a site's capacity to receive and distribute; vulnerability tier (Section 4.1) measures need. The two are read together but never combined into a single ranking.

5.6 Confidence levels and data quality

Given the constraints on open-source data for Cuba, every record carries a confidence rating reflecting the reliability of its location verification, assigned at data entry and reviewed each update cycle.

Table 4. Confidence-Level Criteria

6. Verification and Update Cycle

All records carry a last-verified timestamp and source tier designation, and a data age indicator is surfaced in the interface so users can assess currency before acting. Because domains change at different rates, they are not verified on a single cycle: static assets — facility locations, road networks, ports and airports — are re-verified on the standard cycle, while time-varying conditions such as outage hours, road passability, prices, and connectivity interruptions carry their own observation dates and go stale far sooner. A record whose asset is current but whose condition measure is out of date is displayed as partially stale, not as verified.

7. Limitations

• Non-standardized addressing: Cuban addresses are not consistently standardized across provinces, complicating precise geolocation. Where ambiguity exists, the lower confidence rating is assigned.
• Infrastructure degradation: Cuba's physical infrastructure is deteriorating at a documented and accelerating rate. Condition ratings reflect the most recent available satellite assessment and source testimony but may not capture post-assessment changes; treat them as approximations subject to revision.
• Population data currency: The most recent census was 2022. Emigration since 2021 — estimated at 300,000-500,000 individuals — means municipality figures may overstate current resident populations, particularly in high-emigration provinces such as Holguín and Granma.
• Limited external validation: Scarce open-source data restricts independent verification of many records. The confidence rating makes this uncertainty explicit but does not eliminate it.
• Uneven layer population: Layer maturity varies widely and is stated per layer in Section 5. Social and Infrastructure carry populated layers; Military and Economic are entirely in development; Information has no standalone layer scoped. Several Infrastructure layers have a defined data structure with few or no records. Cross-domain analysis must not read a sparse layer as an absence of conditions on the ground.
• Source constraints on event data: Protest event data updates only after an event, requires an account, and carries no protest detail; state response data is available at moderate OSINT coverage that varies by location and incident. Neither supports real-time situational awareness.
• Incomplete national coverage of security layers: Military installations and internal security facilities are assessed as moderate OSINT availability, with comprehensive national coverage judged unlikely. Absence of a record is not evidence of absence of a facility.
• Sub-municipal resolution: Conditions vary within municipalities while the coding unit is the whole municipality. Derived population figures are upper bounds, and intra-municipal isolation is under-represented.
• Non-attributable sources: Several capacity and condition assessments rest on verbal communication with in-country sources whose identities cannot be disclosed. These are marked by source tier but cannot be independently verified by users.

8. Expected Value

By integrating PMESII-P information in one geospatial environment,Cuba Watch supports:

• Identification of aid-relevant assets, institutions, infrastructure, and population concentrations across each PMESII-P domain;
• Assessment of dependencies across governance, communications, security, and economic systems affecting access and aid delivery;
• Identification of emerging pressure points, vulnerabilities, road/infrastructure status, and areas requiring closer monitoring;
• Analysis of changing conditions over time as they affect the ability to reach people in need; and
• A comprehensive, continuously updated operational picture centered on getting aid to Cuba's people.