Outdoor parking lots are among the most operationally demanding environments for EV charging. Installations must occur in tight, active traffic zones, where bulky enclosures, heavy construction, and complex wiring are impractical. Deployment windows are short, often limited to off-peak hours, and real-world power and network conditions are rarely ideal.

The second barrier is usability. Drivers are less likely to start charging when the process feels fragmented: multiple apps, separate accounts, prepaid cards, and slow, location-specific authorization flows. Additionally, when connectivity is unreliable, cloud-dependent designs can introduce delays and operational risk, resulting in queuing and blocked bays during peak hours.

• Ultra-compact edge AI system optimized for integration into a bollard enclosure.
• NVIDIA Jetson Orin Nano™ delivering up to 67 TOPS of AI performance to support real-time perception and decision making at the edge.
• Single GbE PoE interface to power and connect PoE peripherals over one cable.
• Wireless auto-recovery to re-establish connectivity and maintain continuous operation.
• Outdoor-capable, 24/7-ready design to ensure reliable unattended operation in real-world parking environments.

Axiomtek proposed AIE100-ONA as the edge-computing core for the EV charging bollard, combining high on-device AI performance with a deployment-friendly industrial design.

Bollard-ready form factor: AIE100-ONA is optimized for tight installations, simplifying internal layout and cabling inside the bollard.

Real-time intelligence on-site: AIE100-ONA is built on the NVIDIA Jetson Orin Nano platform, delivering up to 67 TOPS to run local perception and decision logic at the point of service, avoiding cloud round-trip times and maintaining predictable response time.

Simplified peripheral integration: Dual Ethernet includes a dedicated 15W GbE PoE PSE port to power and connect PoE devices over a single cable, reducing wiring complexity inside a bollard.

Resilient connectivity: Smart wireless auto-recovery function automatically re-establishes the wireless connection, supporting near-zero-downtime operation.

Fanless rugged enclosure: Wide temperature support for 24/7 use, with an optional IP42 cover for semi-outdoor operation.

On the AIE100-ONA platform, NVIDIA Jetson Orin Nano delivers the on-device AI compute required to sustain real-time inference with predictable response times at the point of service. Rather than treating the bollard as a simple data collector, the system processes camera and sensor inputs at the source and converts them into real-time operational decisions. This edge pipeline supports license plate recognition (LPR) for vehicle identification, parking-duration tracking to enforce session timing and billing rules, and real-time occupancy classification (occupied/free) to determine stall availability and status. By running these functions locally, the bollard avoids streaming raw video continuously and limits upstream traffic to high-value outputs such as events, state changes, and metadata. The result is a tighter control loop at the curb, lower bandwidth and backhaul requirements, and more consistent operation regardless of cloud processing variability.

• NVIDIA Jetson Orin Nano with Super Mode (up to 67 TOPS)
• Versatile I/O interfaces for peripheral and device integration
• Smart Wireless Auto-Recovery for the zero-downtime operation
• One 15W GbE PoE for camera connectivity
• Wide operating temperature range: –25°C to +70°C

From early evaluation through real-world deployment, Axiomtek supports customers every step of the way, helping them bring edge AI projects into the field smoothly and with confidence. With a broad portfolio of edge AI platforms, customers can choose the system that best fits their performance, form-factor, and deployment needs, backed by hands-on engineering guidance to simplify integration and scale over time.

"We used the Axiomtek systems powered by NVIDIA Jetson in our self-built smart parking bollards. We appreciate their support during implementation, ensuring seamless integration and reliable functionality. This improved our build efficiency and reduced maintenance and management costs by more than 30%. Axiomtek products perform well, and its technical team gives us confidence in our partnership for cloud integration and long-term maintenance," said Dr. Zhen Yi Shon, CEO of VINVERITAS International Corp.