dgaiot

Industrial IoT Aggregation Engine

Erlang/OTP · Million-Device · Shadow · Ontology · gen_statem

---

DLAS Architecture

  ┌──────────────────────────────────────────────────────────────────┐
  │                    EDGE: iotStudio (Python+Vue)                   │
  │  DeviceAccess(9) → UnifiedPipeline → StreamEngine(15) → Alert(6) │
  └──────────────────────────┬───────────────────────────────────────┘
                             │ MQTT / HTTP
  ┌──────────────────────────▼───────────────────────────────────────┐
  │  SECURITY    auth · role · ACL/CLP · Hooks · JWT                 │
  ├──────────────────────────────────────────────────────────────────┤
  │  ACTION      Shadow(gen_statem) · Bridge · MQTT · Rule Engine   │
  │              init → auth → online → {normal, alarm, offline}    │
  ├──────────────────────────────────────────────────────────────────┤
  │  LOGIC       Ontology Engine · Model Registry · 3 ETS Tables     │
  │              load_model → compile → spawn → evaluate → reason    │
  ├──────────────────────────────────────────────────────────────────┤
  │  DATA        Parse(23 Classes) · PG(:7432)                       │
  │              TDengine(_{ProductId}, devaddr NCHAR(50))           │
  │              EMQX(:1883) · Mnesia/ETS                             │
  └──────────────────────────────────────────────────────────────────┘

Full HTML Diagram

FDE Pipeline

Model → Ontology → Device Access → TimeSeries → Rules → Dashboard
  1        2           3              4           5          6

Data Flow

Modbus 40300 → MQTT(dgiot/{site}/{gw}/{dev}/{pt}/data) → Shadow PID
  → evaluate(Rules) → state transition → Parse + TDengine INSERT

Ontology: DLAS

Layer	Function	Key Technology
Data	23 Parse classes, TDengine, Mnesia/ETS	JSONB, SuperTable/SubTable, _{ProductId}
Logic	load_model, spawn_instance, registry, rules, reasoner	3 ETS tables, gen_statem compile
Action	Shadow 1:1 device process, Bridge, MQTT	gen_statem OTP, ont_push_point
Security	auth, role tree, ACL(object), CLP(class), Hooks	JWT, Parse ACL

Full Ontology Doc · Storage Architecture · Station Simulation

Modules

App	Layer	Role
dgiot	Data	Core — EMQX, rule engine, alarms
dgiot_ontology	Logic	Model registry, reasoner, instance spawner
dgiot_parse	Data	Parse Server REST client
dgiot_task	Action	Shadow gen_statem worker
dgiot_device	Logic	Device management + thing model
dgiot_bridge	Action	Protocol bridge framework
dgiot_dlink	Action	Data link layer
dgiot_api	Security	REST API + auth gateway
dgiot_http	—	HTTP service layer
dgiot_tdengine	Data	TDengine timeseries connector

Quick Start

One-click deploy with Docker (recommended for evaluation):

git clone https://github.com/dgiot/dgiot.git
cd dgiot
docker compose up -d

Build from source — the bootstrap script installs Erlang/OTP 24.3 and build tools automatically (Ubuntu / Debian / CentOS / openEuler / macOS):

git clone https://github.com/dgiot/dgiot.git
cd dgiot
bash scripts/bootstrap.sh
make

Documentation

• DGIOT vs. Palantir — positioning: the brain and the limbs, complementary by design
• Open-Source Palantir Landscape — reality-checked survey of "open-source Palantir" projects
• Pragmatic Enhancement Architecture — Kafka + TimescaleDB + rules + BI in four phases (中文版: zh/pragmatic-enhancement-architecture.md)

Related

• iotStudio — Edge Agent & Application (Python + Vue)
• dgiot — Full Erlang IoT Platform

License

Apache 2.0