SOLAR TODO delivered a city-scale Smart Traffic System for Buenos Aires, Argentina, deploying 19 intersections with 4-in-1 Smart Traffic Poles designed for high-precision perception, fast edge decisioning, and standards-aligned traffic communications.
Answer Capsule: In Buenos Aires, SOLAR TODO installed 19 intersections with 4-in-1 smart traffic poles (4K AI + 77GHz radar + adaptive signals). Edge AI on NVIDIA Jetson and 5G/fiber backhaul powered TrafficGPT for real-time optimization.
Background: Traffic pressure meets distributed infrastructure constraints
Buenos Aires faces a dense, multi-modal street network where intersections experience fluctuating volumes, frequent turning movements, and complex vehicle behaviors. In practice, traffic operators must respond quickly to incidents, reduce congestion spillback, and maintain safety under variable demand—often with limited tolerance for installation disruption and with heterogeneous field conditions (lane layouts, mounting constraints, and existing signal cabinet integration).
A second challenge is operational continuity. Intersections require reliable detection and signal control with consistent latency and robust communications to a central decision platform. When latency or detection quality degrades, adaptive signal strategies lose effectiveness and wrong-way behavior becomes harder to detect early.
SOLAR TODO’s approach for Buenos Aires was to standardize a field-ready sensing and control unit—the 4-in-1 Smart Traffic Pole—and connect it to a TrafficGPT central platform through 5G/fiber backhaul using a cooperation model designed to accelerate deployment.
Solution Overview: What SOLAR TODO deployed in Buenos Aires
The project covered 19 intersections, each equipped with 10m L-arm steel poles (dark grey, hot-dip galvanized) configured as a 4-in-1 smart traffic pole:
- 4K AI camera for perception, delivering 98% accuracy, 45+ detection types, and <50ms response.
- 77GHz mmWave radar to improve detection robustness across lighting and weather conditions.
- LED fill light to support consistent visual capture.
- LED signal light for adaptive traffic signaling.
At the edge, each pole runs Edge AI on NVIDIA Jetson, enabling a 5-layer architecture: Perception → Edge AI → Communication (5G/fiber) → City Brain (TrafficGPT) → Applications.
The system was configured with:
- Full 45-type detection
- Adaptive signal control
- Emergency vehicle priority
- Wrong-way alert
The cooperation model: BOT (zero upfront) helped the city accelerate rollout without upfront capex burden, while keeping the technical system fully standards-based.
Finally, the deployment aligned with NTCIP and GB 25280—key requirements for traffic system interoperability and performance expectations.
Technical Specifications
Below are the key technical specifications used in the Buenos Aires deployment (as configured by SOLAR TODO):
- Deployment scope: 19 intersections
- Pole hardware: 19 × 10m L-arm steel pole (dark grey, hot-dip galvanized)
- 4-in-1 Smart Traffic Pole components:
- 4K AI camera: 98% accuracy, 45+ detection types, <50ms response
- 77GHz mmWave radar
- LED fill light
- LED signal light
- Edge AI platform: NVIDIA Jetson
- Edge intelligence capabilities: full 45-type detection, adaptive signal, emergency vehicle priority, wrong-way alert
- Backhaul & central platform: 5G/fiber to TrafficGPT central platform using natural language queries
- 5-layer architecture: Perception → Edge AI → Communication(5G/fiber) → City Brain(TrafficGPT) → Applications
- Standards compliance: NTCIP, GB 25280
- Cooperation model: BOT (zero upfront)
Deployment in Buenos Aires: engineering for field reality
SOLAR TODO’s installation plan for Buenos Aires was built around minimizing disruption at active intersections and ensuring consistent sensing geometry across the network.
1) Intersection-by-intersection pole configuration
Each intersection received a 10m L-arm steel pole in a dark grey finish with hot-dip galvanization, selected to support long-term corrosion resistance in a coastal urban climate. The L-arm geometry provides stable mounting for the camera and radar unit alignment, enabling consistent detection zones for:
- Lane-level object tracking
- Vehicle classification aligned with the configured 45-type detection set
- Wrong-way monitoring in approach directions
2) Edge intelligence to reduce decision latency
Instead of relying solely on backhaul round trips, perception outputs are processed at the edge via NVIDIA Jetson. This architecture supports the project’s low-latency requirements (the camera subsystem provides <50ms response) and enables fast activation of:
- Adaptive signal logic
- Emergency vehicle priority triggers
- Wrong-way alert conditions
This matters in Buenos Aires because intersections can experience sudden demand shifts (commuter peaks, event-driven surges, and incident-driven rerouting). Edge processing helps maintain stable signal behavior even when network conditions fluctuate.
3) Communication to TrafficGPT via 5G/fiber backhaul
For city-level intelligence, SOLAR TODO connected each pole to the central platform using 5G/fiber. The City Brain (TrafficGPT) layer supports natural language queries, enabling operations teams to request and review system states and performance patterns without needing deep system command-line access.
In practice, this supports faster operational workflows:
- Reviewing detection and signal adaptation behavior by intersection
- Investigating alert events (wrong-way, emergency priority activations)
- Coordinating multi-intersection responses through the city brain layer
4) Standards alignment for interoperability and control
The deployment followed NTCIP and GB 25280, ensuring the system is compatible with typical traffic management expectations and control interfaces. This is particularly relevant in Buenos Aires where traffic infrastructure must coexist with existing signal operations and operational procedures.
How the Smart Traffic System improves safety and flow
Adaptive signal control with robust sensing
The 4K AI camera (98% accuracy, 45+ detection types, <50ms response) is complemented by 77GHz mmWave radar. Together, they improve detection reliability across real-world conditions—night lighting variability, changing weather, and vehicle occlusion.
The system’s adaptive signal feature uses these perception inputs to adjust traffic behavior in near real time, reducing unnecessary phase time and improving throughput.
Emergency vehicle priority
When emergency vehicles are detected, the system can apply emergency vehicle priority logic to reduce response time through the corridor. This is achieved through the edge-to-central architecture: edge AI detects and classifies the event, and the City Brain layer coordinates the broader operational context.
Wrong-way alert
The wrong-way alert function addresses a key safety risk in dense urban intersections. By monitoring approach direction consistency and vehicle movement patterns, the system flags suspicious behavior early—supporting faster operator intervention.
Results and Impact
SOLAR TODO’s Buenos Aires deployment delivered measurable operational and safety improvements through a standardized, high-performance sensing and control stack.
Quantified outcomes
- 19 intersections equipped with 4-in-1 smart traffic poles, providing city-scale coverage for adaptive control and safety alerts.
- <50ms response perception capability from the 4K AI camera subsystem, enabling faster signal reactions.
- 98% accuracy visual detection performance and full 45-type detection, expanding the range of traffic scenarios the system can recognize.
- 5G/fiber backhaul connectivity to TrafficGPT, enabling centralized coordination and operations workflows using natural language queries.
Operational impact
- Reduced reaction time for incident-related maneuvers through emergency vehicle priority.
- Improved safety monitoring via wrong-way alert at the intersection level.
- More consistent adaptive signal behavior because edge intelligence (NVIDIA Jetson) performs immediate processing rather than waiting for round trips.
Standards and references (authority for design alignment)
The system’s design and interoperability approach aligns with recognized international and industry frameworks:
- ITU-T guidance for IMT and communications system performance considerations relevant to 5G backhaul architectures.
- IEEE references on intelligent transportation and sensing/communications integration patterns.
- IEC standards relevant to safety and reliability expectations for electrical and communication systems in public infrastructure.
- World Bank and city mobility guidance emphasizing data-driven traffic management and operational resilience.
- Local compliance targets referenced by the project configuration: NTCIP and GB 25280.
Product page and contact
If you want to see how SOLAR TODO designs Smart Traffic System deployments for different intersection patterns, visit our smart-traffic product page. For a deployment discussion in Buenos Aires or other LATAM cities, contact us.
Frequently Asked Questions
Q1: What detection capabilities does the Buenos Aires Smart Traffic System support?
The deployed poles use a 4K AI camera configured for full 45-type detection (45+ detection types) with 98% accuracy and <50ms response, supported by 77GHz mmWave radar for robustness.
Q2: How does the system handle real-time control decisions?
Perception outputs are processed at the edge using NVIDIA Jetson (Edge AI). This supports fast activation of adaptive signal, emergency vehicle priority, and wrong-way alert, while the 5G/fiber backhaul connects to TrafficGPT for city-level coordination.
Q3: What communications architecture was used for the central platform?
Each intersection communicates to the TrafficGPT central platform via 5G/fiber backhaul, enabling city brain workflows and natural language queries for operations and analysis.
Q4: Which standards does the deployment follow?
The project configuration follows NTCIP and GB 25280 for traffic system interoperability and compliance.
Conclusion
The Buenos Aires Smart Traffic System deployment demonstrates how SOLAR TODO combines high-precision sensing, fast edge intelligence, and standards-aligned communications in a field-proven package. With 19 intersections equipped using 10m hot-dip galvanized L-arm 4-in-1 smart traffic poles—integrating a 4K AI camera (98% accuracy, 45+ types, <50ms), 77GHz mmWave radar, LED fill light, and LED signal light—the city gained adaptive control and safety features including emergency vehicle priority and wrong-way alert. Connected via 5G/fiber to TrafficGPT and operated through a BOT (zero upfront) cooperation model, the system provides a scalable foundation for next-stage smart mobility improvements.
Equipment Deployed
- 19 × 10m L-arm steel poles (dark grey, hot-dip galvanized) configured for the Smart Traffic System
- 19 × 4-in-1 Smart Traffic Pole sensing and signaling module: 4K AI camera (98% accuracy, 45+ detection types, <50ms response) + 77GHz mmWave radar + LED fill light + LED signal light
- 19 × Edge AI compute: NVIDIA Jetson (Perception → Edge AI workflow for adaptive signal, emergency vehicle priority, wrong-way alert)
- 5G/fiber backhaul links from field poles to TrafficGPT central platform (City Brain layer for natural language queries)
- Traffic software configuration supporting full 45-type detection, adaptive signal, emergency vehicle priority, and wrong-way alert; standards alignment with NTCIP and GB 25280
- BOT cooperation model for zero upfront deployment