Data Flow
A data flow is a naming system plus a state machine plus a maturity map. All three are required before a device becomes an economic actor.
Naming System
Who named each thing shapes who captures the value.
| Term | Who Coined It | What the Name Hides | Power Concentration |
|---|---|---|---|
| "Smart device" | Consumer electronics | Whether it signs, earns, or can be owned by its user | OEM |
| "Hotspot" | Helium | That it is a token-mining node with routing duty | Protocol operator |
| "Interrogator" | RFID industry | That it is a notary of physical events | Warehouse or integrator |
| "Gateway" | Telecom | That it is a signing authority and time witness | Telco |
| "Digital twin" | Industrial software | Whether the twin is cryptographically bound | Enterprise software vendor |
| "Sensor network" | Academic IoT | Whether readings are signed or tradeable | Platform aggregator |
| "Machine identity" | Peaq / DePIN | That a DID is the machine's wallet address | Protocol |
The naming gap: the industry does not have a shared term for "signed event from an attested reader." Every vendor calls it something different. Until that term is agreed, audits stay manual.
Data Model
Entities, relationships, state transitions.
| Entity | Relationships | State Transitions |
|---|---|---|
| Asset | 1:N tags, 1:N sensors, 1:1 custodian at a time | Created → custodied → transferred → retired |
| Tag | 1:1 asset, N:1 family, readable by many interrogators | Issued → activated → read → revoked |
| Sensor | 1:1 asset or 1:1 environment, emits readings | Commissioned → streaming → drifted → replaced |
| Interrogator | N:1 operator, deployed at 1 location, reads N tags | Deployed → operating → tampered → retired |
| Edge Signer | 1:1 interrogator or gateway, holds a machine DID | Provisioned → signing → compromised → rotated |
| Event | 1:1 signer, 1:N tags/readings, timestamped | Captured → signed → anchored → settled |
| Operator | N:1 network, owns N devices, earns per event | Onboarded → active → slashed → paid |
| Settlement | 1:N events, triggered by a smart contract | Pending → executed → finalised |
Data Footprint
Maturity across the five layers for each major domain. 1 = schema only, 5 = closed feedback loop.
| Data Domain | Schema | Data | API | UI | Feedback | Moat Signal |
|---|---|---|---|---|---|---|
| Asset identity (RFID) | 4 | 4 | 3 | 3 | 2 | GS1 schema exists, settlement thin |
| Environmental sensing | 3 | 4 | 3 | 3 | 3 | Many vendors, no common API |
| Custody events | 2 | 2 | 1 | 1 | 1 | Almost no standard — greenfield |
| Machine identity (DID) | 4 | 3 | 3 | 2 | 2 | Peaq leads, adoption early |
| Positioning (RTK) | 4 | 4 | 4 | 4 | 3 | GEODNET mature, feedback improving |
| Settlement proofs | 3 | 3 | 3 | 2 | 2 | Chain-side solid, UI side weak |
Where moats live: Custody events (score 1.4 avg) is the widest opportunity. Nobody owns the schema for "signed custody handoff." First mover wins.
Decisions Data Drives
Good data turns tacit judgements into automated decisions.
| Data | Decision | Action | Impact of Bad Data |
|---|---|---|---|
| Temperature log | Accept or reject a wine shipment | Release payment, or trigger claim | Healthy cargo rejected, spoiled accepted |
| Custody chain | Authenticate provenance | Mint NFT, or refuse listing | Counterfeit enters secondary market |
| Device attestation | Trust a reading | Accept event, or quarantine | Oracle manipulation, systemic fraud |
| Event timestamp | Order handoffs across borders | Settle cross-border duty | Tax disputes, customs holds |
| Operator uptime | Pay network rewards | Distribute tokens | Free riders, honest operators leave |
Context
- Principles — Parent page and tightness score
- Data Footprint — Commissioning instrument
- Naming Standards — The nomenclature discipline
Questions
Which data domain, if standardised tomorrow, would unlock the most venture opportunity?
- Why does custody events sit at 1.4/5 when every enterprise wants it?
- What would a GS1-equivalent standard for signed custody handoffs look like?