Deploy Plate Recognizer Stream (ALPR) on ZEDEDA

Introduction

Deploying AI-powered computer vision workloads at the edge has traditionally been complex, involving manual configuration, networking, and scaling challenges. ZEDEDA removes this complexity by providing a secure, cloud-managed orchestration platform that allows you to deploy, manage, and scale applications across distributed edge environments. When combined with Plate Recognizer’s advanced license plate recognition (ALPR) software, organizations gain a fast, reliable, and production-ready solution for real-time vehicle analytics at the edge.

ZEDEDA delivers an industry-leading edge orchestration platform that provides centralized visibility, zero-touch provisioning, and lifecycle management for any edge application or device. By leveraging ZEDEDA’s open and vendor-neutral architecture, organizations can deploy workloads seamlessly across x86, ARM, and GPU-based edge nodes without being locked into proprietary infrastructure. ZEDEDA ensures secure and consistent operations from the cloud to the farthest edge—offering granular control, role-based access, and encrypted communication channels for all connected assets.

Plate Recognizer Stream brings AI-powered license plate recognition directly to the edge. It enables real-time detection and analytics of vehicle license plates, make, model, color, and state—even under challenging conditions such as motion blur, low light, or camera angle. This makes it ideal for use cases in transportation, logistics, smart parking, tolling, and security monitoring.

Together, ZEDEDA and Plate Recognizer create a frictionless experience: from onboarding hardware and deploying the ALPR container to monitoring performance—all achievable through ZEDEDA’s intuitive web console or API. With this integration, customers can operationalize computer vision at scale, reduce deployment time from weeks to minutes, and ensure enterprise-grade reliability at every edge location.

Prerequisites 

• You have onboarded your edge node. 
• The edge node must be running a supported version of EVE-OS.
• Outbound HTTPS (port 443) connectivity is required.
• You have either the SysManager or SysAdmin role in your ZEDEDA Cloud enterprise.
• After onboarding is complete, your edge node will appear as Active under Edge Nodes > Devices in the ZEDEDA dashboard.

Create and Attach a Persistent Volume

The Plate Recognizer Stream app continuously processes image frames and metadata. To ensure cached images and logs persist through container restarts or updates, create a persistent volume.

1. Go to Library > Volume Instance > Add New

2. Fill in the following items:

   Attribute	Value
   Name	alpr_volume_1
   Type	Block Storage
   Size	10 GB
   Access Mode	Read/Write
   Encryption	Yes
   Edge Node	onlogic-fr201
   Project	PR-Demo
   Label	alpr1

3. Click Add.

Clone and Validate the App Configuration

Go to Marketplace > Edge Apps > Global Edge Apps to validate, import, or clone the configuration. 

Configure Network Access

Plate Recognizer requires properly configured inbound and outbound network rules for communication with RTSP cameras and its API endpoints. RTSP cameras and their API endpoints are configured on the Plate Recognizer site.

Outbound Rules:

Inbound Rules:

Port 8080 exposure is not required for Stream. If you want to monitor the service, you can expose port 8001 and access it through http://localhost:8001/status/.

Add Cloud-Init Configuration

To inject your Plate Recognizer license key and token securely at launch, use ZEDEDA’s built-in cloud-init system.

Plate Recognizer Stream requires a valid license key. You can obtain a free trial license by registering on the Plate Recognizer website. Visit Stream Live-Camera ALPR | Plate Recognizer ALPR and sign up to receive your trial key via email. After you have your license key, include it in the cloud-init configuration as follows.

Example:

#cloud-config
runcmd:
  - LICENSE_KEY=
  - TOKEN=

Configuration Settings:

• Name: cloud-init

• Variable Delimiter: ###
   

Deploy and Validate

Now you’re ready to deploy the Plate Recognizer Stream container to your device.

1. Go to Marketplace > Local Edge Apps > Plate Recognizer.

2. Click Deploy.

3. Choose Project: PR-Demo, Edge Node: your onboarded device, Deployment Type: Container

4. Attach volume alpr_volume_1 → mount /user-data

5. Apply cloud-init configuration

6. Verify resource allocation: 2 vCPU / 4 GB RAM

7. Click Deploy

Validation:

Open Edge App Instances > Logs to check for the following message:

Test connectivity via RTSP camera stream or http://<device-ip>:8001/status/

Hardware Sizing Recommendations

The performance and number of concurrent camera feeds depend on your device’s CPU, GPU, and available memory. Refer to Plate Recognizer’s official hardware sizing guide for system requirements: Stream Installation Guide | Plate Recognizer.

References & Support

Plate Recognizer Stream: Stream Live-Camera ALPR | Plate Recognizer ALPR

Plate Recognizer Docs: https://guides.platerecognizer.com/docs/stream/getting-started/ 

ZEDEDA Tenant: https://zedcontrol.gmwtus.zededa.net

Support Email: support@platerecognizer.com

Conclusion

By following this article, you can deploy and operate Plate Recognizer Stream (ALPR) on ZEDEDA with persistent storage, secured credentials, and optimal performance monitoring. ZEDEDA’s zero-touch orchestration, combined with Plate Recognizer’s computer vision intelligence, enables scalable and efficient edge deployments that deliver real business outcomes—from reducing operational costs to improving situational awareness and safety. For additional customization or scaling guidance, consult the previous links or contact your ZEDEDA or Plate Recognizer representative.