IoT Fleet Management: Monitoring 200 Industrial Routers Without a Dedicated Team

When your IoT fleet exceeds 50 sites, manual management becomes impossible. Here’s how to design a fleet monitoring system that scales—with two technicians, not twenty.

The Problem with 200 Routers

With 10 routers, we manage them manually. We know which site has which equipment, we maintain an Excel spreadsheet of the configurations, and we conduct regular on-site visits.

With 200 routers, it's a different story:

• Configuration: 200 different configurations, each with site-specific requirements
• Deployment: a new site every 2–3 days during peak activity
• Monitoring: It’s impossible to monitor 200 dashboards manually
• Incidents: Outages occur at night and on weekends
• Updates: Coordinate firmware updates across 200 devices without causing outages

Our client, Faure Industries (an industrial systems integrator with 200 client sites), has solved this problem. Here is their approach.

A 4-Pillar Management Architecture

1. Zero-Touch Provisioning — no manual configuration required

ZTP eliminates the need for manual, site-by-site configuration. Here is the process:

Result: An untrained network technician can deploy an entire site in less than 5 minutes—just plug it in.

2. Deployment Profiles — Centralized Configuration

Faure Industries has created 8 deployment profiles covering 95% of their configurations:

# Profil "Station-Pompage-Standard"
name: Station-Pompage-Standard
vpn: openvpn
modbus:
  baud_rate: 9600
  slaves:
    - slave_id: 1
      polling: 30s
      registers: [pump_status, pressure, flow, level]
mqtt:
  broker: mqtt.eziwan.com
  topic_prefix: "eau/{site_id}/pompe"
alerts:
  - type: device_offline
    delay: 5min
    channel: [email, sms]
  - type: modbus_value_out_of_range
    tag: pressure
    min: 0.5
    max: 8.0
    channel: [email]

When a new site matches this profile, simply assign it to the ZTP token. The entire configuration is deployed automatically.

3. Proactive Monitoring — Anticipating Outages

Reactive monitoring (alerts when a failure occurs) isn't enough for 200 routers. Faure Industries uses predictive monitoring:

Early Warning Indicators

Signal RSRP LTE :
  > -80 dBm    → Excellent (vert)
  -80 -95    → Good (light green)
  -95 at -105   → Degradation (orange) → preventive alert
  < -105       → Critique (rouge) → intervention

Preventive alerts allow you to plan ahead:

• Schedule an antenna inspection before a failure occurs
• Contact the service provider to report a decline in coverage
• Plan to install an external antenna during the next service visit

Fleet Health Dashboard

Active SIM2 (failover): 4 routers · Last outage > 24 hours: 0 routers.

4. Remote Maintenance

Most incidents can be resolved remotely:

Incident	Remote Resolution	Rate
Router crash	OTA reboot via dashboard	85%
Incorrect configuration	Profile redeployment via dashboard	90%
Outdated firmware	Batch OTA update	100%
Modbus slave not responding	Test from dashboard + logs	70%

Only hardware-related issues (power supply failure, antenna disconnected, RS485 cable unplugged) require an on-site visit.

Integration with Existing Monitoring Tools

The value of a fleet management platform is measured by its ability to integrate with the integrator’s or end customer’s existing ecosystem.

Webhooks and REST APIs

Eziwan provides a comprehensive REST API and webhooks for every fleet event:

# Example: Listing offline devices via the REST API
curl -H "Authorization: Bearer $EZIWAN_TOKEN" \
  https://api.eziwan.com/v2/devices?status=offline&site_id=*

# Réponse JSON
{
  "devices": [
    {
      "id": "gw-lyon-04",
      "site": "Lyon-Industrie-Nord",
      "status": "offline",
      "last_seen": "2025-07-01T14:32:11Z",
      "offline_duration_min": 47,
      "sim_active": null,
      "rsrp_last": -108
    }
  ],
  "total_offline": 3,
  "total_devices": 200
}

Available webhooks: device.offline, device.online, failover.triggered, failover.resolved, signal.degraded, firmware.update.completed, certificate.expiring.

Grafana / InfluxDB / Prometheus Integration

For teams that already monitor their infrastructure with Grafana:

# Example of an Eziwan data source in Grafana
datasource:
  type: prometheus
  url: https://api.eziwan.com/metrics
  auth:
    type: bearer
    token: ${EZIWAN_TOKEN}

# Available Metrics
# eziwan_device_rsrp{device_id, site_id, operator}
# eziwan_device_status{device_id, site_id}  # 0=offline, 1=degraded, 2=online
# eziwan_failover_count{device_id, sim_slot}
# eziwan_modbus_errors_total{device_id, slave_id}
# eziwan_mqtt_messages_total{device_id, topic}

A standard Grafana dashboard imported from the Eziwan community displays the following in 5 minutes: a map of sites (online/offline/degraded), a histogram of failover events, and 30-day RSRP signal trends.

ITSM Integration (ServiceNow, Jira Service Management)

For organizations that manage their incidents using an ITSM tool:

// Webhook Eziwan → ServiceNow
{
  "event": "device.offline",
  "device_id": "gw-lyon-04",
  "site": "Lyon-Industrie-Nord",
  "offline_since": "2025-07-01T14:32:11Z",
  "recommended_action": "Check LTE coverage and power supply"
}

This webhook can automatically create a P2-level ticket in ServiceNow or an issue in Jira SM—the on-call engineer is notified without any manual intervention.

Faure Industries Operational Metrics

After 18 months of operation (200 sites):

Metric	Value
Average deployment time per site	4.5 min (excluding shipping)
Collection availability	99.4%
Incidents resolved remotely	83%
Technicians dedicated to fleet management	1.5 FTE
Sites managed per FTE	130

FAQ

Can routers from different brands be managed from a single Eziwan console? Eziwan's centralized fleet management is built into Eziwan gateways. For third-party devices (Teltonika, Cradlepoint, Sierra Wireless), partial integration is possible via the Eziwan REST API or NETCONF/YANG protocols—please contact our team for specific use cases.

How can you quickly detect when a site is offline in a fleet of 200 sites? The Eziwan dashboard displays the status of each site in real time (ONLINE / OFFLINE / DOWNTIME) along with the timestamp of the last communication. Configurable alerts (email, SMS, webhook) are triggered if a site has not reported in for N minutes. For 200 sites, the fleet view (table or map) allows you to immediately identify sites with issues.

Does the OTA firmware update interrupt data collection? On Eziwan gateways, the OTA update is designed to minimize disruption: the firmware is downloaded in the background, and the reboot is scheduled during a configured maintenance window (e.g., 2:00 a.m.–4:00 a.m.). The store-and-forward mechanism ensures that no data collected before the reboot is lost.

---

Further Reading

• Blog: Zero-Touch Provisioning — Deploy 50 Routers in 2 Hours
• Blog: Remote Industrial Monitoring — Complete Guide 2026
• Blog: Industrial 4G Router vs. Consumer Router
• Docs: Eziwan Fleet Management

---

Conclusion

Managing a large-scale industrial IoT fleet isn't a human resources issue—it's an architectural one. With ZTP, centralized profiles, predictive alerts, and OTA maintenance, two people can manage 200 sites just as easily as 20.

The key: eliminating repetitive manual tasks with smart tools, and focusing on activities that truly add value.

Explore the ZTP documentation → · See our use cases →

---

Additional Resources

• Router Fleet Management Guide — methods and tools for monitoring your fleet
• Industrial Routers — catalog of fleet management-compatible devices
• Mass ZTP Deployment Guide — Zero-Touch Provisioning at scale
• Industrial 4G Router — devices with built-in ZTP and OTA management