IoT Sensor Network — 9,000 Sensors, Flood Warning & Environmental Monitoring Across Rio

Updated March 2026

Rio de Janeiro’s Internet of Things sensor network is one of the most extensive municipal IoT deployments in Latin America. With 9,000 georeferenced sensors distributed across the metropolitan area, the system provides continuous monitoring of environmental conditions, traffic flow, infrastructure integrity, and public safety indicators. The network breaks down into two primary categories: 4,000 solid-waste sensors embedded in the city’s culvert and drainage system to detect blockages that could cause flooding, and 5,000 traffic signal sensors for real-time vehicle flow monitoring. Supplementary sensor layers include rain gauges, weather radar, GPS tracking on 10,000 vehicles, and 5,000 planned WiFi access points that double as IoT communication hubs. All data feeds into the Centro de Operacoes e Resiliencia (COR), where 500 professionals on 24-hour shifts integrate sensor intelligence with camera feeds, citizen reports, and social media monitoring to maintain continuous situational awareness across the city.

The Flood Warning Imperative

Rio de Janeiro’s geography makes it one of the most flood-vulnerable major cities in the world. The city is wedged between the Atlantic Ocean, Guanabara Bay, coastal lagoons, and the steep granite peaks of the Tijuca massif and surrounding mountain ranges. Rainfall on the steep hillsides runs off rapidly into narrow valleys and urban drainage channels, creating flash flood conditions that can develop from dry streets to waist-deep water in less than an hour. The catastrophic rains of April 2010 that killed scores of residents and led directly to the founding of COR demonstrated the lethal consequences of inadequate flood monitoring and warning.

The 4,000 solid-waste sensors embedded in culverts across the city represent Rio’s primary defense against the drainage blockages that transform heavy rain from a manageable engineering problem into a life-threatening flood event. Rio’s drainage infrastructure was designed for theoretical water volumes, but in practice, solid waste — plastic bags, bottles, construction debris, vegetation — accumulates in culverts and reduces their carrying capacity. A culvert that is 50 percent blocked by debris will overflow at half the rainfall intensity that it was designed to handle.

These sensors use a combination of ultrasonic level detection, flow rate measurement, and obstruction detection to monitor the condition of each culvert in real time. When a sensor detects rising water levels that deviate from the expected response to current rainfall, or when flow rates drop below thresholds indicating partial blockage, alerts are automatically generated and pushed to COR operators. Maintenance crews can then be dispatched to clear the blockage before water backs up to street level, converting what would have been a flooding incident into a routine maintenance event.

The culvert sensor data integrates with COR’s rain gauge network and weather radar feeds to create a multi-layered flood prediction system. Rain gauges measure actual precipitation at specific locations, weather radar provides spatial coverage showing where rain is falling and where it is moving, and culvert sensors measure the drainage system’s real-time capacity to handle the water being delivered. Machine learning models trained on historical data from all three sources generate neighborhood-level flood risk assessments that COR operators use to issue warnings, deploy pumping equipment, and close flood-prone roads before water reaches dangerous levels.

Traffic Sensor Infrastructure

The 5,000 traffic signal sensors form the second major component of the IoT network, working in concert with the CIVITAS AI traffic system’s 900 AI radars and 3,000 connected traffic signal controllers. These sensors are installed at intersections throughout the city, measuring vehicle counts, queue lengths, turning movements, pedestrian presence, and signal compliance.

Unlike the CIVITAS radars, which are sophisticated AI-powered devices capable of vehicle classification and behavioral analysis, the traffic signal sensors are relatively simple devices optimized for reliability and cost-effectiveness. Inductive loop detectors embedded in the road surface detect the presence and passage of vehicles. Infrared sensors mounted on signal poles measure vehicle counts and pedestrian presence without requiring physical installation in the roadway. Video detection sensors at key intersections provide visual confirmation of traffic conditions that supplements the quantitative data from loop and infrared sensors.

The traffic sensor data feeds into the adaptive signal timing system managed through COR’s Hexagon platform. When sensors at an intersection detect that queue lengths on one approach are significantly longer than on competing approaches, the system can adjust signal timing in real time to provide additional green time for the congested approach. This adaptive capability is particularly valuable during non-recurring congestion events — accidents, construction, special events, weather impacts — where pre-programmed signal timing plans based on historical patterns no longer match actual conditions.

The traffic sensor network also provides the baseline data that powers COR’s predictive traffic models. By analyzing sensor data across thousands of intersections over extended periods, the system identifies normal traffic patterns for every intersection by time of day, day of week, and season. Deviations from these normal patterns trigger alerts that COR operators investigate, potentially identifying incidents or developing congestion before they become severe enough to be visible on camera feeds or reported by citizens through 1746.

The WiFi-IoT Convergence Layer

The planned deployment of 5,000 WiFi access points across Rio serves a dual purpose that extends beyond public internet access. Each access point functions as an IoT communication hub, capable of receiving and relaying data from nearby sensors that communicate via WiFi, Bluetooth Low Energy (BLE), LoRaWAN, or other short-range wireless protocols. This architecture creates a city-wide IoT backbone without requiring dedicated communication infrastructure for each sensor.

The 5,000 access points, each supporting 200 simultaneous users, create a mesh network with a total capacity of one million simultaneous connections. While the primary user-facing function is providing free WiFi to residents and visitors, the IoT functionality operates transparently in the background, collecting data from environmental sensors, parking occupancy detectors, waste bin level sensors, and other IoT devices deployed within each access point’s coverage radius.

This WiFi-IoT convergence layer reduces the cost of adding new sensor types to the network. Instead of building dedicated communication infrastructure for each new sensor deployment, the city can install sensors that communicate through the nearest WiFi access point, leveraging existing power and communication infrastructure. This approach is particularly valuable for pilot projects and experimental sensor deployments where the cost of dedicated infrastructure would be prohibitive for unproven applications.

Environmental Monitoring

Beyond flood warning and traffic management, the IoT sensor network supports environmental monitoring across multiple dimensions. Air quality sensors deployed at key locations measure particulate matter (PM2.5 and PM10), nitrogen dioxide, ozone, carbon monoxide, and other pollutants that affect public health. Temperature and humidity sensors distributed across the city create microclimate maps that inform urban planning decisions about tree planting, park placement, and building orientation.

The environmental monitoring data feeds into the DATA.RIO open data portal, making it available to researchers, health authorities, and the public. University partnerships — particularly with UFRJ’s environmental engineering programs — leverage this data for research on urban heat islands, air quality trends, and the health impacts of environmental conditions on different populations. PUC-Rio’s data science programs use the sensor data to develop predictive models for air quality forecasting that could eventually be integrated into COR’s operational decision-making.

The Porto Maravilha redevelopment area, covering 5 million square meters, represents a showcase for integrated environmental IoT. The area’s modern drainage systems, smart lighting through the Luz Maravilha PPP, VLT light rail integration, and IoT-enabled waste sensors create a district-scale demonstration of what comprehensive urban IoT can achieve. Lessons learned from Porto Maravilha’s sensor deployments inform the broader city-wide rollout, with successful sensor configurations and communication architectures replicated in other neighborhoods.

GPS Vehicle Tracking: The Mobile Sensor Layer

The GPS tracking of 10,000 vehicles — including buses, taxis, metro rail cars, and municipal fleet vehicles — creates a mobile sensor layer that provides traffic intelligence across the entire road network. Unlike fixed sensors at intersections, GPS-tracked vehicles provide continuous speed, heading, and position data along their entire routes, creating a real-time map of travel speeds and congestion across every road that these vehicles traverse.

The bus fleet is particularly valuable as a mobile sensor platform because buses follow fixed routes on regular schedules, providing consistent coverage of major corridors. When bus speeds drop below normal thresholds on a route segment, the system infers congestion even in areas without fixed traffic sensors. This inference capability is validated against CIVITAS radar data and fixed sensor data where coverage overlaps, and has proven reliable enough to generate automatic congestion alerts in areas monitored only by GPS tracking.

Taxi GPS data adds coverage of areas that bus routes do not reach, particularly residential neighborhoods and areas outside the formal transit network. The aggregated movement patterns of thousands of taxis create a heat map of urban mobility demand that informs both transit planning and traffic management decisions. Areas with high taxi demand but poor transit coverage may be candidates for new bus routes or transit service extensions.

Data Architecture and Processing

The 9,000 sensors, 10,000 GPS-tracked vehicles, 5,000 WiFi access points, and supplementary data sources collectively generate a data volume measured in terabytes per day. Processing this data in near real-time requires a computing infrastructure that matches the sensor network in scale and reliability. The COR data center, with 84 servers and nearly 10 petabytes of storage, handles the current data load, while the Rio AI City hyperscale campus will provide the computing capacity needed as sensor density and data resolution continue to increase.

The data processing pipeline follows a three-stage architecture. Edge processing at or near the sensor reduces raw data to structured events — a culvert sensor generates a “water level above threshold” event rather than streaming continuous level readings. These events flow through the communication network to COR’s data ingestion layer, which validates, timestamps, and routes each event to the appropriate processing module. The Hexagon platform then integrates events from multiple sensors into the unified operational picture displayed on COR’s 125-screen video wall.

Historical data from the sensor network accumulates in a data warehouse that supports offline analytics, machine learning model training, and long-term trend analysis. This historical archive is what makes COR’s predictive capabilities possible — the flood prediction models, traffic pattern recognition, and anomaly detection algorithms all depend on years of historical data to establish the baselines against which real-time conditions are compared.

Scaling the Network: What Comes Next

The current 9,000-sensor deployment represents the foundation of a network that will continue to grow as new use cases are validated and additional sensor types become cost-effective. The 5G infrastructure being piloted through the TIM/Enel X/Leonardo MOU will enable denser sensor deployments by providing the connectivity bandwidth and device density capabilities that current networks cannot support. 5G’s massive machine-type communication specification supports up to one million devices per square kilometer, a density that would allow sensor placement at every street corner, every bus stop, and every utility pole in the city.

Emerging sensor technologies that could be integrated into Rio’s IoT network include structural health monitoring sensors for bridges and buildings, noise level sensors for acoustic environment management, water quality sensors for the city’s rivers, lagoons, and Guanabara Bay, and soil moisture sensors for landslide risk monitoring on the hillsides where many of Rio’s favela communities are located. Each of these sensor types addresses specific urban challenges that Rio faces, and the existing IoT backbone — communication infrastructure, data processing pipeline, and COR integration — reduces the marginal cost of each new deployment.

The convergence of expanded IoT sensing with the smart energy grid creates opportunities for energy-aware sensor network management. Solar-powered sensors with battery backup can operate independently of the electrical grid, enabling deployment in locations where grid power is unreliable or unavailable. Smart grid data about local energy availability can inform sensor data collection schedules, reducing transmission frequency during periods of energy scarcity while maintaining full-resolution monitoring during periods of abundant renewable generation.

External Resources

• COR Expansion — Sensor Network Details — Official documentation of the 9,000-sensor deployment
• IDB Smart City Report — Rio de Janeiro — International assessment of Rio’s IoT infrastructure