HEAVY VEHICLE / OFF-HIGHWAY · CANONICAL DATA MODEL

Cummins canonical data model: connect your Cummins equipment to any AI agent

2,461 confirmed mappings·12 protocols·99% field coverage

Forge turns raw Cummins tags into one clean, governed schema your AI agent already understands. One MCP call, no per-vehicle integration project.

Try it free: 3 machines, no credit card → See it work →

Paste your Cummins data and see it normalized in 30 seconds. No POC required.

Read-only. Forge reads your Cummins data through standard industrial protocols. Read-only. Your equipment is never at risk.

The problem

Why Cummins data is hard to integrate

Cummins engines report over J1939 and INSITE. The same oil pressure and turbo RPM readings mean nothing to a fleet AI until the SPNs are decoded and named.

Typical Cummins equipment communicates over J1939 (CAN), HTTP / REST. Each firmware revision, model line and integrator names the same physical quantity differently, so every AI or analytics project pays the integration tax again, the tax that cancels 40% of industrial AI deployments before they ship.

The solution

One universal schema for every Cummins vehicle

Forge resolves Cummins telemetry into Forge's canonical field set, the same 1,027-field canonical schema (366 resolvable from vendor tags today), used across all 47 vendor packs. Point Forge at the vehicle, pass oem: "cummins", and spindle speed is spindle_speed_rpm whether it came from a Cummins controller or anything else on the floor.

J1939 (CAN)HTTP / REST

99%Field coverage

11Canonical heavy vehicle / off-highway fields

3,444Corpus mappings

Field coverage

What Forge resolves for Cummins

Confirmed canonical fields your agent gets from Cummins vehicles, drawn from Forge's 3,444-mapping corpus:

vehicle.engine.oil_pressurevehicle.engine.oil_tempvehicle.engine.fuel_rail_pressurevehicle.engine.intake_manifold_pressurevehicle.turbocharger.rpmvehicle.transmission.oil_tempvehicle.brake.air_pressurevehicle.brake.primary_air_pressurevehicle.wheel_speed_flvehicle.tire_pressure_flvehicle.dtc_description

Sample normalization

Raw tag in		Canonical field out	Meaning
SPN_100: 52	→	vehicle.engine.oil_pressure: 52	oil pressure
SPN_175: 102	→	vehicle.engine.oil_temp: 102	oil temp
SPN_102: 28	→	vehicle.engine.intake_manifold_pressure: 28	boost pressure
SPN_103: 81000	→	vehicle.turbocharger.rpm: 81000	turbo rpm

30-second deployment

Add Cummins to your agent with one call

No SDK, no per-tag mapping, no historian project. Connect the MCP server and normalize live Cummins data:

claude mcp add --transport http forge https://mcp.foundrynet.io/mcp --header 'Authorization: Bearer YOUR_API_KEY'

# normalize raw Cummins telemetry into the universal schema

curl -s https://forge.foundrynet.io/v1/normalize \

-H 'Content-Type: application/json' \

-d '{"data": {"SPN_100": 52, "SPN_175": 102, "SPN_102": 28}, "oem": "cummins"}'

{"vehicle.engine.oil_pressure": 52, "vehicle.engine.oil_temp": 102, "vehicle.engine.intake_manifold_pressure": 28, "coverage": "100%", "confidence": 0.97}

Unknown tags never get force-mapped. Fields below the confidence floor abstain rather than guess, so your agent never acts on a bad reading.

Start free: Forge API, 3 machines → See the full demo →