Chat with your data.
ARDIs deep pool of knowledge – covering not only the assets that make up your system, but the connections between them – make an extremely powerful source of data for LLM-based agents.
Using the industry-standard Model Context Protocol (MCP), AI agents can connect to query all of your ARDI assets, their live values and their history, giving AI the ability to…
Read real-world data
Ask the system what is happening now from real world sensor values, or read from history to analyse behaviour in the past.
Understand context
The AI can easily get the full context of an asset – not only a machine speed, but the temperature, vibration, current product and other related details.
Build reports
Create custom reports and visualisations by simply describing what you want – using our tools and documentation, your model can build web-based live and historical displays.
Search the process
AI models can search your process to trace connections, follow relationships and find assets based on limited or domain-specific information.
Suggest logic & solutions
AI models can not only be helpful in detecting issues that normal analytics might miss, but it can suggest possible solutions , courses of action and prepare documentation.
Safe Access
Have confidence in your integration, knowing that the connection is purely read-only and the model is not capable of writing data to your control system.
Build your own tools
ARDI is built with data-driven, web-based solutions in mind. It has simple APIs which are accessible from browsers, Python and Javascript, meaning that you can easily use tools like Claude Code to create custom reports, charts, widgets and tools.
And if you’ve made something great, you can also publish your custom solutions through ARDIs web interface, giving people across the organisation easy access to your new creations.
Have an Idea
Need a particular chart, a new report, or have an idea for an interactive tool that is specific to you or your industry?
Connect
Point your model (such as Claude Code) to our API documentation. This is built for AI, to give them the information they need to build tools and visualisations.
Get Building
Have a conversation with your model Describe what you’re looking for and collaboratively build up a solution that suits the specific needs of your business and process.
Make it available
Deploy your Python, HTML/Javascript or Electron applications through ARDIs web interface or make calculated values available via OPC-UA.
Unlock the power of Machine Learning.
ARDI also is a great fit for building and deploying machine learning models. ML models are able to run along-side your process, to perform a range of tasks such as…
Anomaly detection
Detect when parts of the system are behaving ‘differently’ – highlighting a wide range of problems before they cause critical defects or downtime.
Predictive results
Find process and quality problems before you even start running your equipment by using predictive AI to let you know what results will ahead of time.
Optimisation
Find the ideal combination of setpoints and targets to maximise your production rate while minimising energy use, waste and defects.
Improved setup
Understand how to setup machines even in extremely challenging environments by factoring in large numbers of different process and environmental conditions.