Version 6.2
Installation
Install ReliVision on a single machine or across a distributed setup.
All-In-One-Machine Self Installation
- System Requirements:
- Windows OS 10/11/Server (via wsl)
- GPU (for AI Training & Inference): NVIDIA GPU with 8GB+ RAM (e.g. RTX3070)
- GPU (for GenAI based Synthetic Data Generation): NVIDIA GPU with 16GB+ RAM (e.g. RTX4080)
- HDD: 250GB + Data Storage (as required)
- RAM: 32 GB+
- Ports Used: 4012, 4013, 4014, 5002(sftp), 5004(gateway), 5432(DB), 5555(http), 5556(https), 5672(message broker), 15672(message broker)
- Prerequisites:
- the following prerequisites should be installed for optimum performance:
- Windows Subsystem for Linux (WSL2) ---> to install open Command Prompt and type wsl - -install
- Docker Desktop for Windows (must be up and running during ReliVision installation and use)
- NVIDIA Drivers for optimum GPU performance
- Installation:
- Client App (ReliUI):
- Click and run the ReliVisionClientApp Installer .exe
- Server App (ReliTrainer & ReliAudit) :
- Click and run the ReliVisionServerApp Installer .exe
- Initate installation by clicking the Starter
- Choose which components (or all) to install from the menu
- Update & Upgrade:
- Client App (ReliUI):
- Click and run the latest ClientApp Installer .exe [existing directory where the Client App (ReliUI) is located should be selected]
- Server App (ReliTrainer & ReliAudit) :
- Click and run the latest ServerApp Installer .exe [existing directory where ReliVision Server App is located should be selected]
- Initiate update or upgrade by clicking the Starter
- Choose which components (or all) to install from the menu
- Server App components are ONLY accessible via ReliUI and are not activated without a valid license KEY. ReliUI prompts the user to enter a license KEY in case the Server App installation lacks it.
Distributed Installation
ReliVision supports distributed deployments across Windows 10/11, Windows Server 2025, and Ubuntu 20.04 or later, allowing components to run on separate machines based on operational requirements. ReliUI is currently available on Windows only.
ReliVision's distributed architecture allows organizations to centralize AI training while running inspections locally across production lines and sites. This reduces hardware costs, simplifies maintenance and enables inspection operations to scale independently.
Training resources can be shared across multiple inspection stations, while edge systems continue operating independently, even without a connection to the central training server.
Example deployment configurations:
ReliVision's distributed architecture allows organizations to centralize AI training while running inspections locally across production lines and sites. This reduces hardware costs, simplifies maintenance and enables inspection operations to scale independently.
Training resources can be shared across multiple inspection stations, while edge systems continue operating independently, even without a connection to the central training server.
Example deployment configurations:
- Edge deployment: ReliAudit and ReliWeb run on remote or disconnected edge machines, enabling local inspection without continuous connectivity to a central server.
- Fully distributed deployment: ReliUI, ReliTrainer, and ReliAudit (including ReliWeb) run on separate machines, allowing training, inspection, and user interaction to be managed independently.
Distributed deployments require proper network configuration to ensure reliable communication between ReliVision components. Relimetrics provides professional support for installation and distributed system configuration.
For assistance or additional information, contact info@relimetrics.com.
For assistance or additional information, contact info@relimetrics.com.
