Manufacturing environments are among the most demanding IoT deployment contexts that exist. The combination of RF-attenuating steel and concrete construction, high electromagnetic interference from variable-frequency drives and high-current switching equipment, extreme temperature cycling in certain process environments, continuous mechanical vibration from production machinery, and the operational consequence of device failure creates a hardware selection environment where consumer-grade and commercial-grade IoT devices underperform.
Device selection for manufacturing IoT deployments is a multi-variable decision that spans physical environmental requirements, data profile characteristics, connectivity technology selection, and the integration requirements of the industrial automation ecosystem the devices need to work within. Getting any one of those variables wrong produces devices that fail early, connect unreliably, or generate data that the manufacturing platform can’t consume. Getting the connectivity variable wrong is common, because connectivity decisions are often made after hardware selection has already locked in the device form factor and radio technology, leaving the connectivity architecture to adapt to choices that were made without it.
This piece covers the device selection framework for manufacturing IoT deployments, the connectivity requirements that different device categories impose, and the infrastructure decisions that determine whether the deployment integrates well with the manufacturing environment it is meant to instrument.
Environmental Requirements Come First
The first filter in manufacturing IoT device selection is the physical environment the device needs to survive, and this filter eliminates a larger portion of the available device ecosystem than most deployment teams anticipate.
Ingress protection rating is the starting specification for any device installed in a manufacturing environment. The IEC 60529 IP rating system describes protection against solid particle ingress (the first digit) and liquid ingress (the second digit). Manufacturing environments that involve machining fluids, wash-down cleaning procedures, high-humidity process conditions, or outdoor exposure typically require IP65 as a minimum: complete dust protection and protection against low-pressure water jets from any direction. Food and beverage manufacturing with regular high-pressure wash-down requires IP67 or IP69K. Devices specified at IP54 or below, which describes the majority of commercial IoT devices, should not be deployed in manufacturing environments without enclosure protection that brings the effective IP rating to the required level.
Operating temperature range eliminates a second category of devices. The standard commercial temperature range for electronics is 0°C to +70°C. Manufacturing environments frequently exceed this range: foundries, heat treatment facilities, industrial ovens, outdoor equipment in cold climates, and refrigerated food processing environments all have operating temperatures outside the commercial range. Industrial-grade IoT devices rated for the extended temperature range of -40°C to +85°C or -40°C to +105°C are required for these environments. Deploying commercial-grade devices in out-of-range temperature environments produces early failure rates that can’t be attributed to any single fault mode but are consistently correlated with the thermal exceedance.
Vibration and shock resistance affects any device mounted on or near production machinery. Continuous vibration from motors, presses, conveyors, and compressors degrades solder joints, connector contacts, and plug-in SIM tray connections over time in ways that are not visible externally and don’t trigger immediate failure. The failure mode is intermittent connectivity that becomes progressively worse over months, which is difficult to diagnose because the device passes functional tests when stationary. Devices intended for high-vibration mounting should use MFF2 soldered SIM form factor, conformal-coated PCBs, and vibration-rated connector specifications.
EMI tolerance in manufacturing environments with significant electromagnetic interference sources requires attention to device shielding specifications. Variable-frequency drives, induction heaters, arc welding equipment, and large motor starters all generate electromagnetic interference that can affect radio performance and sensor accuracy. Device selection should include EMI immunity ratings per IEC 61000-4 series standards relevant to the installation environment.
Device Categories and Their Connectivity Requirements
Manufacturing IoT deployments span several device categories with distinct connectivity requirements, and the category-by-category assessment determines the connectivity architecture for the full deployment.
Condition monitoring sensors on production machinery, including vibration sensors for bearing fault detection, acoustic emission sensors for cutting tool wear, and current transformers for motor health monitoring, have connectivity requirements driven by the data volume and latency profile of the monitoring application. As covered in previous technical work on sensor connectivity, this category divides into:
- High-frequency raw waveform capture requiring 4G Cat-4 or 5G NR for direct cloud transmission, or edge compute with FFT processing before transmission over LTE-M
- Processed spectral feature transmission that fits within LTE-M bandwidth when edge inference runs locally
- Summary statistics and threshold-based alerting that can operate over NB-IoT for the lowest-frequency monitoring applications
Environmental and process monitoring sensors covering temperature, humidity, pressure, flow, and level in process environments have data profiles well-matched to LPWAN technologies. NB-IoT or LTE-M with PSM configuration delivers the multi-year battery life that sensors in locations without convenient power access require, and the small, infrequent payloads of environmental monitoring fit within LPWAN bandwidth constraints. The physical environment requirements for this category are often the most stringent: process temperature sensors may need to operate at temperatures that require remote-mounted electronics with thermally isolated probe connections.
Production line quality control systems using computer vision for defect detection, dimensional verification, or assembly confirmation are among the highest-bandwidth IoT applications in manufacturing. A camera-based quality station capturing images at production line speeds, which may be tens of parts per minute in high-speed manufacturing, generates image data volumes that require local edge processing before any cellular transmission is practical. The connectivity requirement for production quality systems is typically a high-bandwidth connection from the edge compute node to the manufacturing execution system (MES) or cloud platform, where 5G NR or wired Ethernet is the appropriate architecture rather than LPWAN cellular.
Asset and tooling trackers on high-value tools, fixtures, work-in-progress assemblies, and returnable packaging within a manufacturing facility have connectivity requirements shaped by the indoor location accuracy requirement and the power budget of the form factor. Bluetooth Low Energy with BLE gateway infrastructure aggregating location data is the dominant technology for indoor asset tracking within a facility boundary. For assets that move between facilities or between facility and customer, cellular-based trackers with LTE-M for battery-efficient operation are appropriate for the outdoor and in-transit segments of the asset journey.
Gateway and edge compute devices that aggregate data from multiple sensor categories before transmitting to the cloud platform are the devices with the most demanding connectivity requirements in many manufacturing IoT architectures. The aggregate data volume from a dense sensor population may require 4G Cat-4, 5G NR, or FWA for the gateway-to-cloud backhaul, even when the individual sensors connecting to the gateway operate over NB-IoT or LTE-M. Gateway connectivity selection needs to be based on the aggregate upstream data volume, not on the per-sensor data profile.
OT Protocol Integration and the Industrial Automation Ecosystem
Manufacturing IoT devices do not operate in isolation from the existing industrial automation infrastructure. The PLC, SCADA, DCS, and MES systems that govern manufacturing operations have established communication protocols that IoT devices need to integrate with, and device selection needs to account for this integration requirement.
The dominant OT communication protocols relevant to IoT integration in manufacturing include:
OPC UA (OPC Unified Architecture) is the current standard for secure, platform-independent data exchange in industrial automation. An IoT device or gateway that implements an OPC UA server exposes manufacturing data to OPC UA clients, including MES platforms, historian systems, digital twin engines, and cloud IoT platforms that implement OPC UA clients, without requiring custom protocol translation. Device selection for manufacturing IoT deployments that need to integrate with modern OT infrastructure should prioritize OPC UA support as a connectivity requirement alongside the cellular radio specification.
MQTT with Sparkplug B is the IoT protocol layer that bridges OT and IT environments in many IIoT implementations. Sparkplug B defines a payload specification for MQTT that provides the structured data model, birth/death certificate semantics, and state management that industrial automation integration requires. IoT edge gateways that implement Sparkplug B allow MQTT-based cloud platforms to consume manufacturing data in a format that carries the operational context the OT environment tracks.
Modbus RTU and Modbus TCP remain widely deployed in legacy manufacturing equipment and instrumentation. IoT gateways with Modbus master capability can read registers from PLCs, drives, and instruments that do not have modern communication interfaces and expose that data via OPC UA or MQTT for cloud integration. Device selection for deployments that need to connect legacy equipment requires evaluating the gateway’s protocol library, not just its radio specification.
EtherNet/IP and PROFINET are the dominant industrial Ethernet protocols in their respective ecosystems. Gateways that bridge between these OT protocols and IT-standard IP networking are the integration point for manufacturing IoT deployments in environments built on Rockwell or Siemens automation infrastructure.
The connectivity implication of OT protocol integration is that many manufacturing IoT deployments are better served by a gateway-centric architecture, where field devices use wired industrial Ethernet or serial protocols to connect to a gateway and the gateway provides both OT protocol integration and cellular connectivity for cloud transmission, than by direct cellular connectivity at each sensor. This architecture reduces the number of cellular SIMs in the deployment, simplifies the connectivity management footprint, and concentrates the cellular radio hardware in a gateway device that can be specified for the full environmental and electrical requirements of the installation location.
Private Cellular Networks for Manufacturing
The manufacturing floor is one of the highest-value deployment environments for private LTE and private 5G networks, and device selection for manufacturing IoT needs to account for whether the deployment architecture includes private cellular infrastructure.
Private 5G in manufacturing provides several capabilities that shared public cellular networks cannot:
- Deterministic latency with quality of service guarantees that allow time-sensitive automation applications, including robotic coordination, AGV navigation, and closed-loop control, to rely on cellular connectivity with latency profiles comparable to wired industrial Ethernet
- On-premises data processing where all IoT data stays within the facility network without traversing public carrier infrastructure, satisfying operational technology security requirements that prohibit manufacturing data from leaving the plant network
- Dense device support at the sensor density that modern manufacturing instrumentation requires, without the spectrum contention that shared public networks experience in high-device-density environments
- Coverage control where the private network covers the facility footprint precisely without spillage to neighboring facilities or public areas
Device selection for private cellular manufacturing deployments requires devices that support the specific 5G NR band allocated to the private network, typically CBRS in the United States (Citizens Broadband Radio Service, Band 48) or dedicated spectrum in other markets, which is a different specification from public carrier 5G support. Not all commercial 5G IoT devices support CBRS or private network spectrum configurations, and this requirement needs to be part of the hardware selection criteria before devices are procured.
KORE’s connectivity infrastructure supports public cellular and private network architectures for manufacturing IoT deployments. The ConnectivityPro platform manages public network SIMs across 40+ carrier partnerships and 200+ countries for deployments where public cellular is the appropriate connectivity architecture, while KORE’s private network capabilities support manufacturing environments where on-premises 5G is the right choice for the latency, security, and coverage requirements of the operation. With 21M+ active SIMs deployed across industrial and manufacturing verticals, the infrastructure reflects the operational experience of large-scale manufacturing IoT deployments across facility types and geographies.
Device Selection as a System Architecture Decision
Manufacturing IoT device selection is a system architecture decision where the choice of device form factor, environmental rating, radio technology, OT protocol support, and connectivity configuration collectively determine whether the deployment delivers sustained operational value or accumulates maintenance costs and data gaps that erode the business case.
The sequence that produces the best outcomes is consistent across manufacturing IoT deployments that perform well:
- Define the physical environment requirements first, including IP rating, temperature range, vibration exposure, and EMI environment
- Identify the data profile for each device category, including payload size, transmission frequency, latency requirement, and mobility profile
- Select connectivity technology for each category based on the data profile and the connectivity infrastructure available at the deployment site
- Evaluate OT protocol integration requirements and determine whether a gateway-centric or direct-cellular architecture better fits the existing automation infrastructure
- Assess whether private cellular is appropriate for the facility’s latency, security, and coverage requirements, and adjust device specifications accordingly
Getting this sequence right before hardware procurement locks in the device specifications is the difference between a manufacturing IoT deployment that performs as designed across its operational lifetime and one that requires costly remediation after commissioning.
If you are in the design phase of a manufacturing IoT deployment and evaluating the connectivity architecture alongside your hardware selection, talk to a KORE connectivity engineer about the network infrastructure requirements your device population and facility environment demand.
Published June 30, 2026