惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

大猫的无限游戏
大猫的无限游戏
MyScale Blog
MyScale Blog
雷峰网
雷峰网
量子位
小众软件
小众软件
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
博客园 - 叶小钗
T
Tailwind CSS Blog
月光博客
月光博客
博客园 - 【当耐特】
博客园_首页
罗磊的独立博客
博客园 - 三生石上(FineUI控件)
IT之家
IT之家
爱范儿
爱范儿
阮一峰的网络日志
阮一峰的网络日志
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
WordPress大学
WordPress大学
The Cloudflare Blog
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
S
SegmentFault 最新的问题
人人都是产品经理
人人都是产品经理
V
V2EX
酷 壳 – CoolShell
酷 壳 – CoolShell

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
SCADA, PLC, and BMS: a practical field guide to industria...
Atlas Scada · 2026-06-18 · via DEV Community

If you have ever walked past a boiler room, a pumping station, or the rooftop ventilation units of a shopping mall, you have walked past a quiet layer of electronics and software that keeps all of it running. That layer is called industrial automation, and the words people use around it — PLC, SCADA, HMI, BMS, "dispatching" — are not interchangeable. Here is a practical guide to what each one actually does, written from the point of view of the people who build and commission these systems.

The three layers

Almost every project, whether it is a heating plant or an "intelligent building", stacks into three layers.

1. The field layer — instrumentation (KIPiA).
Temperature, pressure, flow, level, CO2, current. Sensors turn physical reality into signals (4-20 mA, 0-10 V, RTDs, or digital buses); actuators do the reverse — valves, dampers, variable-frequency drives, contactors. Get this layer wrong (wrong range, wrong placement, no isolation) and nothing above it can save you.

2. The control layer — PLC / controllers.
A Programmable Logic Controller reads the field, runs the logic, and drives the actuators on a deterministic cycle measured in milliseconds. Typical hardware: Siemens S7, Schneider Modicon, OWEN, or a Carel pCO for HVAC and refrigeration. The PLC is what keeps a pump from running dry or a heat exchanger from freezing — and it has to work even when everything above it is offline. That is the golden rule: safety-critical logic lives in the controller, never in the SCADA.

3. The supervisory layer — SCADA / HMI.
SCADA stands for Supervisory Control And Data Acquisition. It collects data from many controllers, shows operators a live mimic diagram (the HMI), logs trends, raises alarms, and lets an operator change setpoints. It sits above the PLC, it does not replace it. For buildings the same idea wears a different hat: a BMS (Building Management System) is essentially SCADA tuned for HVAC, lighting, energy and access.

How the layers talk

The glue is protocols. In practice you meet three constantly:

  • Modbus (RTU/TCP) — old, simple, everywhere.
  • OPC UA — the modern, secure, vendor-neutral choice for controller-to-SCADA and SCADA-to-IT.
  • BACnet — the lingua franca of building automation.

A real site is almost never single-vendor. The integrator's job is to make a Carel chiller, a Siemens boiler controller and an OWEN panel all show up as clean, named tags on one screen. A gotcha worth knowing: writing a setpoint over Modbus is not merely "write a register" — you must respect the data type and scaling. A float32 setpoint written as a single 16-bit word becomes a meaningless denormal number. Small detail, big outage.

"Dispatching": monitoring that does not sleep

Dispatching is the operational practice layered on top of SCADA: one place from which a small team watches many distant objects 24/7 and reacts before a tenant ever calls. Modern dispatching is usually edge + server:

  • an edge node at each site keeps polling controllers and buffering data even when the internet drops;
  • a central server aggregates every object, pushes alarms to phones, and stores history.

That architecture is what lets one engineer keep an eye on dozens of boiler houses from a laptop.

Three mistakes that separate a toy from a real system

  1. Alarm flooding. Auto-generating a min/max alarm on every tag feels thorough; in reality it buries the operator under hundreds of nuisance alarms. The ISA-18.2 alarm-management standard exists precisely to fight this — alarms must be rationalized, prioritized and actionable.
  2. Stale data shown as live. When a device goes offline, a naive HMI keeps displaying its last value as if it were current. The operator then "sees" 21 C in a room that is actually freezing. A real system flags data as stale based on the poll interval.
  3. Trusting the controller's clock. Schedules drift because the controller's battery-backed clock drifts. More than once we have chased a "heating turns on an hour early" complaint straight to a dying coin-cell battery. A good system cross-checks time against NTP and alarms on drift.

Who builds this

This is the work of system integrators: choosing the instrumentation, programming the controllers, building the SCADA or BMS, and commissioning it on site. At Atlas Scada we design and deliver this end to end — instrumentation and control, PLC programming, SCADA/BMS and round-the-clock dispatching — so that quiet layer stays quiet.

If you are specifying your first automation project, the single best early decision is this: keep critical logic in the controller, insist on a documented tag list, and treat alarms as a design artifact, not an afterthought.