Overview
When you run more than one Raspberry Pi with LoopString, keeping track of every device individually gets difficult fast. Device Groups and the Fleet Overview dashboard solve this by giving you two complementary tools:
Device Groups let you organize your Pis into named, color-coded collections. A group is just a label — devices can belong to one group at a time, and you can rearrange membership at any time without affecting the devices themselves or their sensor data.
Fleet Overview is a dedicated dashboard page that shows all your active devices side by side, with live online/offline status, active alert counts, system health warnings, and quick links to each device's individual dashboard. The page also shows fleet-wide totals: how many devices are currently online, how many active alerts exist across the whole fleet, and the combined sensor and actuator count.
Fleet Overview is available on the Pro plan and above. Device groups are stored in your account, so they are preserved across sessions and devices.
How to Create and Manage Device Groups
Creating a Group
Open the device picker by tapping the device selector at the top of any dashboard page. You will see an option to manage groups alongside your list of devices. Give the group a name that reflects the physical location, project, or team responsible — for example "Greenhouse A", "Building 3", or "Production Line 1". Choose a color to make the group easy to identify at a glance.
Adding Devices to a Group
From the group management screen, select the devices you want to add from your device list. You can add multiple devices to a group in one step. Each device can belong to one group; if you add a device to a second group, it moves out of the first.
Renaming or Recoloring a Group
Tap the edit icon next to any group name to rename it or change its color. Changes take effect immediately and are reflected everywhere the group appears in the interface.
Reordering Groups
Groups have an order number that controls how they are listed. You can drag groups into your preferred order from the group management screen. The order is saved to your account and persists across sessions.
Removing a Device from a Group
Open group management, find the device in the group's member list, and tap the remove button. The device returns to the ungrouped state. Its sensor data, dashboard configuration, and settings are not affected.
Deleting a Group
Deleting a group removes the label and its membership list. The devices that were in the group remain in your account, active and accessible — they simply return to the ungrouped state. No sensor data or device configuration is lost.
The Fleet Overview Dashboard
Fleet Overview is accessible from the bottom navigation bar (the chart icon) or by navigating to Fleet in the main menu. It gives you a single-page summary of your entire deployment.
Summary Tiles
At the top of the page you will find six summary tiles showing fleet-wide counts:
- Online — how many of your active devices are currently connected and reporting data
- Offline — how many active devices have not sent a heartbeat recently
- Critical Alerts — the total number of sensor readings currently breaching a critical threshold across all devices
- Warnings — the total number of active warning-level threshold breaches across all devices
- Total Sensors — the combined count of sensors reporting across all online devices
- Total Actuators — the combined count of actuators configured across all online devices
These numbers update in real time as devices come online, go offline, or trigger new alerts.
Device Cards
Below the summary tiles, every active device gets its own card. Each card shows:
- The device name and a live online/offline status indicator with a pulsing dot when the device is connected
- The time since the device was last seen, when it is offline
- The number of sensors and actuators on the device
- Alert badges: critical alerts appear in red and warnings in amber, both showing the count and linking directly to the alert log for that device
- System health notices: if CPU usage, disk usage, or memory usage on the Pi are above healthy thresholds, a badge appears on the card naming the affected resource and current value
- A "Go to Dashboard" button that takes you directly to that device's live sensor and actuator dashboard
The fleet page subscribes to each device's live state in real time, so the cards reflect what is happening now, not a cached snapshot.
Use Cases
Farm Operations Spanning Multiple Sites
A farm running three greenhouses, each with its own Raspberry Pi, can create a group per greenhouse — "Greenhouse North", "Greenhouse East", "Greenhouse West" — and use Fleet Overview to check plant health across all three from one screen. If a temperature sensor in Greenhouse East triggers a critical alert overnight, the Fleet Overview's critical alert count changes immediately and the badge appears on that device's card. The operator can tap through to the East greenhouse dashboard, review the reading, and decide whether to open a vent or send someone to investigate — all without having to check each greenhouse individually.
The group structure also means that when the operation adds a fourth greenhouse, it takes seconds to create a new group and assign the new Pi to it. Everything else — dashboard setup, alert thresholds, automation rules — is done on the individual device dashboard as normal.
Multi-Site Commercial Deployment
A facilities management company overseeing environmental monitoring at several office buildings can create a device group per building. Fleet Overview gives the monitoring team a heads-up view of all buildings at once, showing which locations have online devices and which have connectivity problems. The online/offline tile at the top of Fleet Overview makes it easy to spot a Pi that has lost its network connection before tenants or building managers notice anything wrong.
When multiple buildings trigger warnings at the same time — for example, during a hot summer day when HVAC systems are stressed — the warnings total in the summary tiles gives the team an immediate sense of scale. They can prioritize which building to address first based on whether the alerts are critical or warning-level.
Troubleshooting
Fleet Overview Shows an Upgrade Prompt Instead of Devices
Fleet Overview requires the Pro plan or higher. If you see an upgrade prompt when you navigate to Fleet, your account is on the Hobby, Maker, or Starter plan. You can upgrade from Settings under Subscription. Device groups are accessible at all paid tiers; the Fleet Overview page itself is the gated feature.
A Device Card Shows as Offline Even Though the Pi is Running
The online/offline status in Fleet Overview is determined by whether the Pi has sent a heartbeat to LoopString within the expected interval. If the Pi is running but the card shows offline, check the following:
First, confirm the Pi has an active internet connection. SSH into the Pi or check its network settings directly. Node-RED needs to reach Firebase to report status.
Second, check whether the Pi's LoopString agent service is running. On the Pi, run the Node-RED process monitor or look at the system logs. If Node-RED stopped unexpectedly, restart it and wait one publish interval (up to 60 seconds on the Hobby plan, faster on paid plans) for the status to update in Fleet Overview.
Third, confirm the device has not been accidentally set as inactive in device settings. Inactive devices do not appear in Fleet Overview.
Alert Counts in the Summary Tiles Do Not Match What I See on the Individual Device Dashboard
Fleet Overview counts only currently active alert breaches — sensor readings that are actively outside a configured threshold at this moment. If you recently acknowledged an alert or the sensor reading returned to a normal range, the count will decrease. The individual device dashboard may show alert history that includes past events that have since resolved. Check the Alarm Thresholds documentation if you need to adjust when alerts trigger or clear.
A Device Card Shows Health Warnings for CPU, Disk, or Memory
Health warning badges on a Fleet device card mean the Pi itself is under resource pressure, separate from sensor readings. A CPU warning, for example, means the Pi's processor usage is above the warning threshold — often caused by a runaway process, a large Node-RED flow, or too many sensors publishing at a high frequency.
To investigate, navigate to that device's dashboard and open the Device Health panel in settings. You will see a real-time graph of CPU, memory, and disk usage. If CPU is consistently high, consider reducing the sensor publish frequency for that device via the Live Meter settings or reviewing the Node-RED flow for any loops or high-frequency inject nodes.
I Deleted a Group but My Devices Disappeared
Deleting a group does not delete or deactivate the devices in it. If you deleted a group and do not see the affected devices, they may have been filtered out of the current view. Check the device picker to confirm the devices are still listed in your account. If a device is genuinely missing, it may have been removed from your account separately — contact support with the device serial number.
Known Issues and Limitations
A device can belong to only one group at a time. If your operational structure requires overlapping groupings — for example, devices grouped by both location and team — the current group system does not support this. Use the device name and dashboard room names to encode additional context.
Fleet Overview does not currently support filtering device cards by group. All active devices are shown together on the page. If you manage a large number of devices, the page may require scrolling to find a specific device.
Group membership is stored in your LoopString account, not on the Pi itself. If you transfer a device to another account, its group membership is not transferred — the new owner will need to assign the device to a group in their account.
The Fleet Overview summary tiles aggregate data across all subscribed devices as they come online. On first load, if a device has not yet streamed its state, its sensors and actuators may not be counted in the totals until the subscription catches up. Waiting a few seconds after the page loads will generally resolve this.