Question 1

Does the gateway support 4–20 mA sensors directly?

Accepted Answer

Yes. The Eziwan Gateway has 2 to 4 4–20 mA analog inputs (depending on the model) directly on the terminal. These inputs are connected via screw terminals on a DIN rail—no external converter is required for standard industrial sensors (PT100 temperature via transmitter, pressure, level, flow). The resolution is 12 bits (4,096 levels), with galvanic isolation to prevent interference from ground loops. Scaling (unit, min/max values for 4 mA and 20 mA) can be configured via the web interface.

Question 2

What is the minimum measurement interval?

Accepted Answer

For 4–20 mA analog inputs and Modbus registers, the minimum polling interval is 100 ms (10 Hz). For most industrial equipment (temperature, pressure, level), a 1-second interval is more than sufficient. For applications requiring a high sampling rate (vibrations, motor current), dedicated data acquisition modules are preferable. On the cloud platform, data is stored at the acquisition rate and can be viewed with a resolution of up to 100 ms.

Question 3

How long is the data stored?

Accepted Answer

The standard retention period is 12 months for all raw data. Data is stored in a time-series database (InfluxDB) with automatic compression after 30 days: data older than one month is aggregated by the minute (average, min, max, standard deviation), and data older than six months is aggregated by the hour. This allows us to retain 12 months of useful historical data while keeping storage footprint under control. An Extended Retention plan allows raw data to be retained for up to 5 years.

Question 4

How can you manage alerts from several hundred devices without getting overwhelmed?

Accepted Answer

Eziwan offers several anti-flooding mechanisms. First, deduplication: an alert generates only one notification, even if it remains active for 8 hours. Second, site-based grouping: alerts from each site can be aggregated into a single summary email rather than individual emails. Configurable timeouts (minimum 5 minutes active before notification) eliminate false alarms caused by transient issues. Finally, quiet hours (nighttime, weekends) allow non-urgent alerts to be filtered out during non-business hours.

Question 5

Can you perform predictive maintenance with Eziwan?

Accepted Answer

Eziwan provides the building blocks of predictive maintenance: continuous high-frequency data, drift detection (values that gradually deviate from the nominal value), rate-of-change (abnormal rate of change), and correlations between variables. For AI/ML algorithms (failure prediction models, anomaly detection), you can export the data via the REST API to your analytics tools (Python, Azure ML, AWS SageMaker). The Eziwan platform does not include a native ML engine but integrates natively with industrial analytics platforms.

Question 6

Is Eziwan compatible with CMMS systems such as SAP PM or Maximo?

Accepted Answer

Yes, via the REST API and webhooks. When an alert is triggered, Eziwan can send a webhook to your CMMS to automatically create a work order. For SAP PM, integration is done via SAP PI/PO or directly via the SAP REST API. For Maximo, the Maximo REST API supports creating work orders via HTTP POST. Native connectors to Mobility Work and the leading cloud-based CMMS platforms are available in the Enterprise plan.

Question 7

How do you connect vibration sensors for monitoring industrial motors?

Accepted Answer

Vibration monitoring of industrial motors relies on two types of sensors, depending on the budget and the required accuracy. ICP (Integrated Circuit Piezoelectric) accelerometers, also known as IEPE, are the standard sensors for spectral analysis: they measure vibrations from 0.5 Hz to 10 kHz with a typical sensitivity of 100 mV/g and excellent accuracy for detecting frequencies characteristic of bearing faults (BPFO, BPFI, BSF, FTF). They connect via a BNC cable and require a bias power supply (a 2–20 mA constant current source provided by the acquisition module). 4–20 mA vibration sensors with analog output (e.g., Wilcoxon 786A, IMI Sensors 682A) are a simpler alternative: they convert the RMS value of the vibrations into a 4–20 mA signal that can be read directly by the Eziwan gateway’s analog inputs without requiring a specific module. Typical alarm thresholds: 2.8 mm/s RMS for a warning, 7.1 mm/s for a critical alarm (ISO 10816-3 standard for motors ranging from 15 kW to 300 kW). The Modbus connection is used for smart vibration analyzers such as the SKF IMx-1 or the IFM VSP001, which directly provide RMS, peak, crest factor, and FFT metrics via the RS-485 Modbus RTU bus. The correlation between vibration and temperature is particularly useful: a simultaneous increase in bearing temperature (+15°C) and vibration (twice the nominal value) is a strong indication of imminent bearing failure.

Question 8

How many sensors can be connected to an Eziwan gateway?

Accepted Answer

The connection capacity of an Eziwan gateway depends on the type of interface used. On the RS-485 bus (Modbus RTU): The Modbus RTU standard supports up to 247 slave addresses on an RS-485 bus, but in practice, the physical limit of an RS-485 segment without a repeater is 32 nodes (32 slaves + 1 master). The Eziwan gateway has 2 independent RS-485 ports, allowing up to 64 Modbus RTU devices to be connected in parallel (32 per port). Each device can expose dozens or hundreds of Modbus registers—for example, an Eastron SDM630 meter exposes 120 variables on a single slave address. Over Ethernet (Modbus TCP): via an industrial Ethernet switch on the same local network as the gateway, the number of Modbus TCP devices is theoretically unlimited—the gateway can poll as many PLCs or Modbus TCP gateways as needed, with the only limits being available bandwidth and the polling cycle time. For 4–20 mA analog inputs: the gateway has 4 direct analog inputs (depending on the model), each capable of measuring a 4–20 mA sensor (pressure, temperature via a transmitter, flow, level). To increase this number, Modbus RTU expansion modules with 8 or 16 analog inputs (such as the Wago 750-455 or Phoenix Contact AXL F AI8 modules) can be connected to the RS-485 bus. In practice, virtually all industrial installations (even large sites) can be covered by a single Eziwan gateway thanks to the combination of RS-485 and Modbus TCP.

Question 9

How do I set up predictive alerts for condition-based maintenance?

Accepted Answer

Eziwan offers three alert levels that cover the progression of an asset’s condition, from basic monitoring to advanced condition-based maintenance. Level 1 — Absolute Threshold Alerts (Basic Rule): Trigger an alert when a variable exceeds or falls below a fixed value. Example: Trigger an alert if the motor bearing temperature exceeds 85°C for more than 5 consecutive minutes. This type of alert is the simplest to configure and covers basic equipment monitoring. Level 2 — Trend-Based Alerts (for predictive maintenance): Eziwan calculates the rate of change of a variable over a configurable sliding window (1 hour, 24 hours, 7 days). Example: trigger an alert if the bearing temperature increases by more than 0.5°C per hour—which signals gradual deterioration even before reaching the absolute threshold. This trend-based detection (rate-of-change) is at the heart of predictive maintenance: it anticipates a failure with several days’ advance notice. Level 3 — Alerts based on the maintenance schedule (conditional maintenance): Schedule reminder alerts based on actual operating time (cumulative motor hours), the calendar date, or a combination of both. Example: Generate a work order in the CMMS after 4,000 hours of compressor operation or every 6 months, whichever comes first. For customers seeking more sophisticated anomaly detection models, the Eziwan REST API allows you to export raw data to analytics platforms (Azure IoT Analytics, AWS SageMaker, Dataiku) that apply ML models (isolated forest, LSTM) and return anomaly scores via webhook to the Eziwan alert engine.

Industrial Monitoring —From measurement to real-time alerts

An IoT monitoring platform
designed for industrial use

Multi-protocol acquisition

Multi-channel alerts

Predictive Maintenance

Automatic PDF Report

Advanced Configurable Thresholds

CMMS / GMAO Integration

From Sensor to Alert in 4 Steps

Connect the RS-485 sensors

Configure Alert Thresholds

Get alerts

Analyze trends

Industrial monitoring:
Three pieces of equipment, three challenges

High-Voltage Transformers

Electric Motors

Tanks and Vessels

Everything you want to know about
monitoring of industrial equipment

Start monitoring your equipment
manufacturers starting today