AI and analytics
AI and analytics for IoT sensor data through MCP
Let an AI model analyse your sensor data. Through an MCP server and the open API it reads structured data from Sensor-Online, finds patterns and deviations, and writes the report.
500+ customers · 100,000+ active sensors · Trusted by Akademiska Hus, Vattenfall, LKAB and HSB
How it works
From sensor data to insight.
Sensor data
Readings are stored in Sensor-Online.
API and MCP server
Structured JSON data is fetched through the open API.
AI model
The model analyses the data, through n8n or directly.
Insight
Comparisons, deviations and reports.
What you can do
Questions that used to need an analyst.
Compare week to week
For example compare energy use with weather conditions from one week to the next.
Detect deviations
Let the system find readings that stand out from the normal pattern.
Automatic reports
Have findings summarised and reported without manual work.
Build on open data
The analysis uses structured data from the open API.
Cloud AI or your own model?
Choose where the analysis runs.
| Cloud AI model | Local LLM | |
|---|---|---|
| Connection | MCP connection to a cloud model, such as Claude, with n8n | MCP connection to an LLM you run yourself |
| Where analysis runs | In the AI provider’s service | In your own environment |
| Best for | Getting started quickly | Organisations that already run a local LLM and want to keep data inside |
A tool for people, not a replacement
Check results before you act.
AI findings are a starting point. Review important deviations and recommendations with someone who knows the operation before acting on them.
Related features
Secure IoT platform with REST API
Read more →Integrate IoT data with business systems
Read more →Node-RED and n8n automation
Read more →Alarm management
Read more →Frequently asked questions
How does AI connect to Sensor-Online?
Through an MCP server that fetches structured JSON data from the open API, and through n8n, so an AI model can analyse it.
What can the AI do with my data?
Compare periods, such as energy use against weather from week to week, detect deviations and produce reports.
Can I use my own AI model?
Yes. If you already run a local LLM, it can be used instead of a cloud model.
Does my data have to leave my environment?
Not with a local LLM. With a cloud model, the data you choose to analyse is sent to that provider’s service.
Is this ready-made or custom?
It is a flexible capability built on the open API. Talk to us about the analyses you want.
Want to put AI to work on your sensor data? Talk to an engineer.
Tell us what you want to find out, and we will show how it can be done.
All trademarks belong to their respective owners.