ICP DAS will showcase an AI-enabled intelligent transportation system at APTA 2026 TRANSform & EXPO in Chicago, combining NVIDIA GPU-based traffic video analytics with its redundant iTLC-9000 traffic signal controller. The system is designed to connect real-time traffic detection, emergency-vehicle recognition and violation monitoring with adaptive signal management at smart intersections.
ICP DAS is taking an AI-and-edge-computing approach to intelligent transportation systems at APTA 2026 TRANSform & EXPO, where it will demonstrate a traffic intersection combining computer vision with redundant signal-control infrastructure.
The Taiwan-based industrial computing and automation company will exhibit at Booth 4625 at McCormick Place in Chicago from October 4–7. APTA describes TRANSform & EXPO as a major public transportation technology showcase, with the 2026 event expected to bring more than 12,000 attendees from 85 countries and more than 800 exhibiting companies.
The centerpiece of ICP DAS’s demonstration is the iTLC-9000 Series Redundant Multifunctional Traffic Signal Control System. The company will use a simulated intersection with two sets of physical traffic signals to demonstrate real-time signal operation.
Rather than treating traffic lights as isolated control devices, the system combines signal control with AI-based video analysis. ICP DAS says its demonstration will use NVIDIA GPU technology to analyze live traffic video, identify traffic-flow conditions, detect approaching emergency vehicles, identify traffic violations and recognize associated license plates.
The resulting information can be fed into the traffic-control layer, allowing the iTLC-9000 to adjust signal timing and support operational responses. The approach reflects a broader movement toward intelligent transportation systems in which cameras, edge computing and control infrastructure operate as a connected system.
The hardware architecture is designed around redundancy. According to ICP DAS, the iTLC-9000 supports up to 16 traffic signal phases and complex intersections with as many as 32 approaches or controlled ways. Its modular design includes hot-swappable modules intended to simplify maintenance without requiring the entire controller to be taken offline.
The controller also provides rapid primary-to-backup switchover, abnormal-status notifications and proactive maintenance functions. These features address a different requirement from AI analytics: ensuring that the physical traffic-control infrastructure continues operating when a component fails.
That combination of resilience and intelligence is important for connected transportation infrastructure. A computer-vision model can identify traffic conditions, but its value depends on the system’s ability to convert those observations into reliable traffic-management actions.
Cybersecurity is another component of the iTLC-9000 architecture. As traffic controllers become increasingly connected to networks, transportation operators face a larger attack surface spanning controllers, cameras, edge computers, communications equipment and centralized management platforms.
The AI component adds another layer of computing at the intersection. Instead of sending every video stream to a distant cloud environment, GPU-equipped edge systems can process traffic data closer to where it is collected. This can reduce dependence on continuous cloud connectivity and potentially shorten the time between detection and a control-system response.
For emergency-vehicle detection, for example, the system can analyze approaching traffic and identify relevant vehicles before passing information to the signal-control layer. The same architecture can be applied to traffic-flow monitoring and violation detection, according to ICP DAS.
The company’s wider demonstration will include the WISE-5231 Edge Computing Controller, PET-7016M Ethernet Remote I/O Module and MDC-714 Modbus Data Concentrator. It will also show the DL-1038-WF Environmental Monitoring Module, ALM-Horn-WF-BR Piezoelectric Alarm Horn and iTLC light-control modules.
Those products illustrate the infrastructure required around an intelligent intersection. Sensors generate information, remote I/O connects field equipment, edge controllers process or route data, communication systems move information between components, and signal controllers ultimately execute traffic-management commands.
This layered architecture is increasingly relevant as public transportation agencies explore connected and automated technologies. APTA’s 2026 EXPO specifically highlights technology, autonomous vehicles, enhanced mobility, safety and security, and zero-emission transportation infrastructure among the trends shaping the industry.
The event’s scale also illustrates why transportation technology vendors are increasingly targeting integrated platforms rather than individual components. APTA says its EXPO is designed to showcase technologies and practices across the public transportation sector, bringing together transit personnel, policymakers, government agencies, manufacturers, suppliers and consultants.
For AI infrastructure providers, intelligent transportation is another example of edge AI moving into physical environments. Traffic intersections generate continuous streams of visual and sensor data, while decisions can have immediate real-world consequences. That makes latency, reliability, cybersecurity and hardware redundancy important alongside model accuracy.
ICP DAS’s approach combines those requirements in a single deployment model: GPU-powered computer vision provides the intelligence layer, edge controllers provide local processing and connectivity, and the redundant iTLC-9000 provides the traffic-signal control layer.
The model also points toward a more autonomous form of intersection management. Instead of relying solely on fixed signal schedules or centralized operator intervention, AI systems can identify changing traffic conditions and feed those observations into automated control processes.
The company is positioning the technology for smart-city applications where traffic efficiency, emergency response and infrastructure reliability need to operate together. Whether such systems can deliver measurable improvements will depend on factors including detection accuracy, integration with existing traffic-management systems and the operational policies governing automated signal changes.
At APTA 2026, however, ICP DAS will demonstrate the underlying architecture directly: AI vision at the edge, connected field devices and a redundant traffic controller working together at a simulated intersection.
That convergence is becoming a defining feature of intelligent transportation technology, as cities move from connected infrastructure toward systems capable of sensing, analyzing and responding to changing conditions in real time.
Market Landscape
Intelligent transportation is moving toward connected systems that combine sensing, edge computing, AI analytics and automated control. APTA’s 2026 event highlights technology, autonomous vehicles, enhanced mobility and safety alongside conventional public transportation infrastructure.
The technical challenge is increasingly one of integration. Computer vision can identify vehicles and traffic conditions, but transportation operators also need reliable controllers, communications, cybersecurity and fallback mechanisms. Edge AI can keep analysis closer to the intersection, while redundant control architectures can maintain operations when individual components fail.
This creates a market that spans industrial automation vendors, transportation technology providers, GPU and edge-computing companies, computer-vision developers and traditional traffic-signal manufacturers. AI-enabled signal control is consequently becoming part of the wider AI Infrastructure and AI Agents & Autonomous Systems conversation, particularly where software is allowed to influence physical infrastructure.
Top Insights
- ICP DAS will demonstrate AI-powered traffic analysis alongside redundant signal control at APTA 2026 TRANSform & EXPO in Chicago.
- NVIDIA GPU technology will support video analysis for traffic flow, emergency-vehicle detection, violations and license-plate recognition.
- The iTLC-9000 uses modular redundancy, hot-swappable components and backup-controller switching for intersection reliability.
- Edge computing can process traffic video closer to intersections, supporting lower-latency analysis without relying exclusively on cloud infrastructure.
- APTA’s 2026 EXPO expects more than 12,000 attendees and 800 exhibiting companies from 85 countries.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
