JasperNode Technical Architecture (Controller Runtime)
Execution Model: JasperNode replaces the traditional, indiscriminate cyclic scan of a PLC. It uses a deterministic, event-driven, four-stage "Main Cycle":
Data In: Reads all values from external hardware/IT systems.
Event Preparation: Identifies only the tags whose values have changed.
Script Execution: Executes *only* the scripts triggered by those specific value changes.
Data Out: Writes all resulting output values in a single batch.
Logic Language: Control logic is written and executed in JavaScript
Core Principle (Atomic Tag Control): A variable (a "tag") can only be changed by its own single, assigned script. This rule eliminates common bugs where multiple program parts write to the same variable.
Core Components: separated from UI and AI Chat in an isolated thread (Worker).
Tag Engine: A persistent, publish/subscribe system that manages all tags, values, and triggers script execution.
Connector Manager: Manages all data I/O to/from OT hardware and IT systems
Script Manager: Manages logic development, execution and versioning.
Version Control: All logic scripts and its changes are stored within a local DB, providing a complete change history and rollback capability.
AI Agent Integration
The primary role of the AI Agent is to translate natural language prompts from a human user into executable JavaScript control logic, create/modify tags and configure the connectors.
Prompt: A user writes a natural language command to generate logic or create connectors into the on-site JasperNode software.
Contextualise and Generate (At JasperX): The prompt is securely sent to JasperX, where it adds essential system context. The AI agent generates all necessary components for the task, including the required connector configurations, new tags (variables), and the JavaScript control logic.
Review & Deploy: The generated code and configurations are sent back to the user. As the "Human-in-the-Loop," the user can review and approve the items (critical tasks) before they are deployed into the live system.
Hardware Connection
A diagram is worth a thousand words.
⚠️ The internet connection is only for programming purposes and not required for the code execution.
Once the software is installed on your PC or Industrial PC, follow these steps to get it running:
Open your web browser and navigate to your node's web interface. This is typically http://localhost:9088 (if you're on the same machine) or http://[your-device-ip]:9088
Link the "node" to JasperX:
At JasperX, create a project and click "Link"
At JasperNode, copy the Node ID from the System Config page
At JasperX, paste the Node ID & send OTP (One Time Password)
At JasperNode, copy the OTP and paste it into the JasperX window.
Click "AI Assistant" at JasperNode web UI and type your first prompt.
JasperNode supports several Industrial and IT Protocols to connect to hardware and your software infrastructure.
Hardware Connectors (OT)
Modbus TCP (Client and Server)
Siemens S7 - to read/write Siemens PLCs
Ethernet/IP (in testing phase)
EtherCAT (in testing phase)
OPC UA Client (in testing phase)
BACnet/IP (in development phase)
Software Connectors (IT)
MQTT - Modern Industry 4.0 protocol (sparkplug B only)
InfluxDB - Database Logging
Emails - Send emails
HTTP Server - Create a mini API server
HTTP Client - read data from websites
Node-RED - read and write tags
The documentation is embedded in the JasperNode. Once installed, open AI chat and ask "How to get started?". The AI assistant will guide you through the Key Concept and the Steps to take.
For a deep dive into our architecture and the technology behind it, download the full technical white paper.
You own 100% of the intellectual property you create.
Any control logic, process configurations, tag structure or proprietary data you develop using JasperNode is your exclusive property. JasperNode is a tool, just like a code editor or a compiler and JasperX is a management platform for your "nodes". The IP belongs to the creator - your company.
Your data is strictly confidential and completely private.
Access to your prompts, your data, and the code you generate is restricted exclusively to the authenticated user. Our platform is built on a secure, multi-tenant architecture. This means we enforce strict logical isolation at both the application and database levels. Our application code and database policies ensure that a query from your account cannot see or access data from any other organisation. No other customer or any unauthorised third party can see your information.
No. Absolutely not.
Your proprietary data, your prompts, and your business processes are never used to train AI models for other customers. Your competitive advantage always remains yours alone.