Platform
Talks to the field you already have.
Your RTUs, PLCs and flow computers already speak standard protocols. The platform connects over those, and turns what comes back into one consistent well and facility model.
No hardware replacement programme. No rip-and-replace step.
- JR14.TK03.LEVEL14.62ftgood
- JR14.MTR.GASRATE318.4MCF/dstale 41m
- JR14.TK01.LEVELno datano comms
The standard protocol stack
Direct device communication over the protocols already deployed across upstream and midstream fields.
- Modbus RTU and Modbus TCP.
- DNP3, including the event-driven reporting it was designed around.
- OPC UA for equipment and systems that expose it.
Publish-subscribe for constrained links
MQTT with Sparkplug B through an edge gateway, for sites on metered or intermittent cellular where polling every register on a cycle is impractical.
- Report-by-exception rather than continuous polling, so a metered link carries changes instead of repetition.
- Sparkplug B birth and death certificates make device state explicit rather than inferred from silence.
- The gateway holds the field-side connection, so a dropped link is a known state rather than missing data.
Tag normalization
Field equipment from different vendors names the same measurement differently. The platform maps those into one model so a reading means the same thing regardless of what produced it.
- Tags resolve to a well, facility or tank in a consistent asset hierarchy.
- Units are captured with the tag and retained, never silently converted.
- The mapping is inspectable — you can see which device tag became which model point.
Connection state is data
A device that has stopped answering is a fact worth surfacing, not a gap to paper over.
- Communication state is reported rather than inferred from an absent value.
- A stale reading carries its age; it is never relabelled as current.
- The platform does not invent an intraday timestamp for a value that did not carry one.
Why normalization is the part that matters
Connecting to a device is the easy half. The half that decides whether a platform is useful in year two is what happens to the tag once it arrives.
A field with equipment from four vendors and fifteen years of installation history has four different names for tubing pressure, three different units conventions, and at least one register that means something entirely different on one make of controller than on another. Left alone, that inconsistency propagates into every chart, every alarm threshold and every report built on top of it.
The platform resolves incoming tags into one well and facility model, retaining the original device tag alongside the model point so the mapping stays inspectable. When a number looks wrong, you can follow it back to the register it came from.
Common questions
Do we need new field hardware?
No. If your equipment speaks Modbus, DNP3 or OPC UA today, the platform connects to it as it is. An edge gateway is only needed where you want MQTT and Sparkplug B over a constrained link.
What if a site has poor cellular coverage?
That is the case the publish-subscribe path exists for. Report-by-exception over MQTT with Sparkplug B carries changes rather than repeatedly polling every register, and device state is explicit rather than guessed from silence.
What about wells with no instrumentation at all?
Those are covered by gauge capture rather than telemetry. A pumper records the tank gauge on a phone and it enters the same pipeline, carrying its own provenance.
See it against your own wells.
A demo runs against synthetic data first, then against a read-only connection to your system of record if you want to see real numbers.