Edge Node
Learns the asset it is mounted on.
Request a pilot unitEDGE AI ENGINE
Powering continuous learning on the device
- Real-time on-device intelligence
- No cloud required
- No pre-training on your data
It adapts to your environment in real time, learning continuously as it runs. A Neuromodulation Layer, inspired by how the brain gates learning, signals when learning should pause, and your team decides — without heavy compute.
LearnDetectAlert
Why RailMind
Neuromorphic, but it learns differently.
| RailMind | Fixed alarm sensors | Cloud models | Other edge AI | |
|---|---|---|---|---|
| InstallLearns each asset's own normal by itself | Yes | No | Partly | No |
| SetupNo labelled fault data needed | Yes | Yes | Partly | Partly |
| MaintenanceKeeps learning on site | Yes | No | Partly | No |
| SecurityData stays on site | Yes | Yes | No | Yes |
| Team controlYour team starts and stops learning | Yes | Not applicable | No | Not applicable |
YesPartlyNoNot applicable
“Partly”: cloud models are usually retrained centrally, in batches, not on the device; some edge AI is trained on normal data only.
Our engine keeps learning on site, and signals when to pause, so your team can keep real damage from being learned as normal wear.
How our Neuromodulation Layer enables continuous learning →The technology
See the technology →Schematic animation.
Learns on the device
The model forms on the device, from the data of the asset it watches.
No training data up front
No labelled faults and no data-collection phase before it is useful.
Keeps learning, under your control
It follows the asset as it ages and its load changes. Your team pauses and resumes learning.
Built for small hardware
Designed from the start for low-power microcontrollers, not shrunk down from a data-centre model.
Static alarms cry wolf or miss the warning, driving up the cost of manual inspections.
Where it applies
All applications →
Industrial equipment
Predictive maintenance (PdM)
For plant and maintenance teams.
Each machine gets a model of its own normal, learned where it runs.
Infrastructure
Structural health & network monitoring (SHM)
For owners and operators of structures and networks.
Each structure is read against its own history, on site.
Vision
Video and image analytics on site
For production quality teams and road operators.
Each camera learns its own scene, where it is mounted.
Inside your devices
Engine licence
For makers of sensors, machines and devices.
The same engine, built into the product you already make.
Talk to us about a licence →The products
All products → 
Edge Hub
Brings a whole site together.
Request a pilot unit
Edge GTS
Same clock, same map.
Request early accessProducts in development. Pilot units on request.
Building your own hardware? License the engine →
Start with a few machines and grow with the results. How it works →
Where we have worked
Settings we have worked in through data analysis, demonstrations or joint projects. Names are withheld.
Questions
Do we need examples of failures to get started? +
No. Each device learns what normal looks like for the asset it sits on. Records of past faults are useful, but not required.
Does it replace our existing monitoring? +
No. It runs alongside what you have. Your existing alarms and inspection rounds stay as they are while you see what it adds.
Can it use the sensors we already have? +
In many cases, yes. An Edge Node can take an external sensor on a cable, using the standard industrial signals and connections common in your region. Cameras plug straight into the Edge Hub. We check your sensors at the start of a pilot.
How are the devices powered and mounted? +
Edge Nodes are available in a battery-powered version and with external power; the Edge Hub and Edge GTS run on external power. A node can be mounted with a magnet, bolts or adhesive, depending on the surface.
How does it connect to our systems? +
Nodes reach the Edge Hub over LoRa radio, in the frequency band permitted in your region; Bluetooth and Wi-Fi are used only to set up and activate a node. The Hub joins your network by cable or Wi-Fi and passes readings to your dashboards, maintenance tools or platform through our API.
How does our team see a change? +
In whichever way suits you: in your own system through our API, in our web interface, in our app, or by email.
What happens after maintenance, or when operating conditions change? +
Your team decides. Learning pauses during the work and resumes when you say so, so the device learns the new normal instead of reporting it as a change.
Where does our data go? +
It stays on the device unless you choose to send readings on. When you do, they go to your own systems.
Are the devices certified? +
Not yet: the devices are still being designed and built. We plan to certify them to the standards of each market we sell in, and to add explosion-protection certification for hazardous areas.
What does a pilot involve? +
You choose the assets: one machine or structure, or a few of the same kind. If you already collect data from them, we first review it with you to check that the site suits and to agree what success looks like. We then bring the devices, install them with your team and connect them to your systems. At the end we go through what the devices showed against your own records, and you decide whether to go further.