Skip to content
MACT.ai
All work

IoT Control System

A connected control platform spanning devices, gateways, apps and cloud.

2026Hardware · Connectivity · Apps · Cloud
IoT Control System

Project Overview

A control system for distributed equipment. Nodes run locally and keep working when the network does not; a gateway aggregates them, and operators supervise the fleet from web and mobile.

Challenge

Sites had unreliable connectivity, mixed hardware generations and safety behaviour that could not depend on the cloud being reachable.

Solution

Control logic lives on the node. The cloud sets intent and observes outcomes, which means a lost uplink degrades reporting rather than control. A versioned device protocol let old and new hardware share one fleet.

Architecture

01

Hardware

Sensor and actuator nodes with isolated I/O, plus a gateway that bridges local buses to the cloud.

  • Isolated industrial I/O with protection
  • RS-485 and CAN local buses
  • Gateway with cellular and Ethernet uplinks
  • DIN-rail enclosure, wide temperature range
02

AI

Anomaly detection on telemetry to surface drift and failure signatures before they escalate.

  • Unsupervised anomaly detection per node
  • Drift detection across the fleet
  • Maintenance signals ranked by confidence
  • Operator feedback loop into the model
03

Software

A versioned MQTT protocol, an offline-first control runtime, and web and mobile consoles for operators.

  • Offline-first control with local safety interlocks
  • Versioned protocol across hardware generations
  • Fleet OTA with per-site staging
  • Role-based access and full audit trail

Result

  • Control unaffected by uplink loss
  • One fleet across three hardware generations
  • Fleet-wide updates without site visits
  • Anomaly signals reached operators before failures

Technology Stack

Hardware

Industrial I/ORS-485CAN

AI

Anomaly DetectionDrift Monitoring

Embedded

RTOSLinux GatewayOTA

Cloud

MQTTTime-seriesWeb + Mobile