June 16, 2026

This is the largest release of the year so far. It adds an AI copilot, a full lifecycle for alarms, custom severities, new ways to define alerts and dashboards at scale, and a set of security features each client can turn on. Several of these features are off by default; each section says how to enable it.

New features

AI Copilot

A conversational assistant in Monitor and Manager that answers questions about the platform's live data.

Access: the floating AI Copilot button on any screen opens a side panel.
What it can look up: active alarms and their details, facilities, endpoints, dashboards, endpoint status, a configuration review and metric explanations. Answers stream in, in the user's language.
Suggestions: when the conversation is empty, it suggests questions based on the actual state (for example, active alarms or offline devices).
Actions with approval: it can propose creating a dashboard or an alert condition. The proposal shows up as a card with Apply / Discard and only runs if the user approves it and holds the required permission.
Options: reasoning level, answer length and model (Auto, Fast or Deep).
Quota: an indicator in the header shows the percentage used.
Privacy: credentials and tokens are stripped from the text before it is sent to the model.
Availability: this is an add-on and is off by default. Available on request; contact support.

Operational context for AI (AI Context)

Describe each client's operation so the copilot gives more relevant answers.

Access: Manager → client settings → AI Context tab. Facility, device, endpoint and user records also have an AI Context tab.
Predefined templates: Apply predefined template offers 10 profiles (NFPA 72 and EN-54 fire detection, industrial warehouse, manufacturing, electrical substation, water treatment plant, cold chain, wildfire prevention, waste management and generic IoT operations). Applying a template replaces the existing context.
Structured forms: for a facility, building type, floor area, fire load, suppression, climate zone, hazardous materials and applicable standards.
Completeness: each level shows a bar with how complete it is.

AI onboarding assistant

A full-screen chat that takes a new user from scratch to a proof of concept with real data.

Access: Monitor → side menu → Onboarding.
Flow: connect (LoRaWAN or an MQTT gateway), receive data and build the dashboard. The assistant creates the device model with its script, registers the devices, explains how to connect the network server or broker, and confirms when data arrives.
Wrap-up: at the end it offers Go to my dashboards or starting over.
Languages: the interface is available in English and Spanish.
Availability: off by default. Can be enabled per client on request; contact support.

Alarm lifecycle

Alarms now have states and owners, and every change is recorded.

Access: Monitor → side menu → Alarms.
States: New → Viewed → In process → Resolved. Each transition is logged with user, date and comment. Resolving requires a comment.
Actions: Mark as viewed, Take in process, Resolve and Reset. Like on a fire panel, reset only works once the condition is back to normal.
Silence: Silence until acknowledged or for 10, 30 or 60 minutes. A silenced alarm does not escalate, and users are reminded when the silence expires.
Escalation: if nobody handles the alarm, the group configured for its severity is notified once by email, SMS, push or WhatsApp.
Auto-resolve: configurable per severity.
View: status, device, endpoint, severity, facility, assignee, opening time and duration, with a toggle between active alarms and history.
Availability: off by default. Can be enabled per client on request; contact support.

Custom alarm severities

Each client can define its own severities on top of the four built-in ones.

Access: Manager → Alarms → Alarm severities.
Settings: multi-language name, color, order, Auto-resolve, Requires comment on resolve and Notify push.
Where they are used: they can be picked in the alert editor and in alarm types, and they show up with their color in alarm lists, on the map and in dashboard counters.
Restrictions: built-in severities are read-only, and a severity in use cannot be deleted.

Rate-of-change alerts

An alert can now fire on how fast a variable changes, not just on its value.

Access: Manager → Alarms → Alerts → create or edit an alert → Rate of change condition.
Settings: window in minutes, threshold per minute and Direction (Both, Rising or Falling).
Composite alerts: the condition can also be used inside AND/OR alerts.
History: in Monitor → Historical data, the Rate of Change (Δ/min) checkbox plots the rate of change instead of the value.

Alert templates

Define an alert once and apply it to many devices.

Access: Manager → Alarms → Alert templates.
Settings: severity, endpoint, variable type, trigger and reset conditions and thresholds, minimum duration and email recipients.
Targets: by device model, by facility or by manual selection.
Modes: Static (Fork) skips devices that already have the alert; Dynamic (Link) updates the alerts it already created when re-applied.
Result: Apply to devices shows how many alerts were created, updated, skipped or failed.
Deletion: if the template is deleted, the alerts it created remain as regular manual alerts.

Per-client security: password policy, authenticator-app 2FA and compliance dashboard

A set of security features enabled per client, which appear in the client's Security menu.

Password policy: minimum length, lowercase, uppercase, numbers and symbols, recovery link validity, password expiration, password history and account lockout after failed attempts. It can only tighten the global policy, never relax it. If a user belongs to several clients, the strictest policy applies.
Two-factor authentication: 2FA with an authenticator app (Google, Microsoft, Authy, 1Password, Bitwarden). Users enroll by scanning a QR code and confirming with a code from the app. At login the app code is requested, with email or SMS as a fallback.
Compliance dashboard: cards with 2FA adoption for administrators and for all accounts, locked accounts, failed attempts in the last 24 h, active sessions, pending password changes and the oldest unrotated password.
2FA policy: 2FA can be required for everyone or only for administrators, and the allowed channels (app, SMS) can be chosen.
Access: Manager → [client] → Security → Password policy / Two-factor authentication / Compliance dashboard.
Availability: off by default. Can be enabled per client on request; contact support.

Inactivity logout

Access: Manager → client settings → Security tab → Inactivity Timeout block.
Settings: per user group: Disabled, 5, 10, 15, 30 or 60 minutes. If a user is in several groups, the shortest timeout wins.
Warning: a modal with a 30-second countdown and Continue session / Log out now buttons. If there are unsaved changes, users get a 5-minute grace period.
Multiple tabs: activity is synchronized across open tabs.
Audit: the logout is recorded in the activity log, and the login page shows why the session ended.
It is disabled for all groups by default.

Log in as another user

Lets an administrator see the platform exactly as a user does, for support purposes.

Access: Manager → [client] → Security → Users → Log in as row action.
Reason: an optional reason can be entered when starting.
During the session: a fixed banner in Monitor and Manager shows that you are acting as another user and lets you return to your own session. The session lasts one hour at most.
Traceability: it is recorded in the activity log under the administrator's real identity, and the user receives an email notice.
Limits: a client administrator can only do this with users of their own clients, and platform administrators cannot be impersonated.

Device migration between facilities

A wizard to move a device to another facility while keeping its history.

Access: Manager → Devices → Migrate button on the row.
Step 1, destination: within the same client (between facilities) or to another client.
Step 2, pre-flight check: endpoints, alerts that will be disconnected, sub-devices that will move with it, previous migrations and active alarms.
Step 3, confirmation: accept the consequences and re-enter your password.
Gateways: devices behind a gateway move with it.
Availability: off by default. Can be enabled per client on request; contact support. Migration to another client is reserved for platform administrators.

Client-wide scheduled reports

A view where the client administrator can see and manage every report subscription, not only their own.

Access: Manager → [client] → Security → Scheduled reports.
Columns: report, owner, recipients and next run, with a filter.
Actions: edit the subscription (recipients, PDF, Excel or PNG format, days and times) or delete it.

Dashboard templates, presets and filter chips

Tools to generate dashboards at scale from rules instead of building them one by one.

Access: Manager → Dashboards section → Templates, Presets and Filter Sets.
Templates: define the layout with rules that repeat per device, per endpoint or per gateway, filtered by model, endpoint type or variable. The dashboard rebuilds itself whenever a device in the facility is created, edited or deleted.
Presets: apply a template to a facility and publish it in the Monitor side menu, with a label, icon and order.
Filter Sets: sets of chips (All, Ideal Situation, Disconnected, by severity or by alarm type) with a visual editor.
In Monitor: dashboards opened from a preset show the chip bar with device counts. Selecting a chip hides the widgets that don't match and compacts the grid.

Advanced map widget

The dashboard map widget now supports several map types.

Access: Monitor → dashboard in edit mode → map widget → parameters.
Types: facilities (the existing one), devices, alarms with a heatmap, asset tracking with historical routes, and custom map.
Detail panel: clicking a marker opens a side panel with device or facility data, live metrics, battery, signal and active alarms.
Severity colors: the pin takes the color of its most severe active alarm.
Map style: every map gets a style picker (standard, high contrast, silver, retro, night, satellite and terrain), remembered per user.

Control widget

A widget to send commands to one or more endpoints from a dashboard.

Access: Monitor → dashboard in edit mode → add the Control widget.
Three-step setup: control type (switch, slider or button) and command; endpoints to control (only compatible ones are listed); visual customization, with an option to ask for confirmation before running.
Security: depending on the endpoint's settings, the widget asks for a password, asks for confirmation, or runs directly.
Read-only: on shared dashboards, without permission, or when the facility is in maintenance.
Audit: every execution is recorded in the activity log.

Hexagonal grid widget

Access: Monitor → dashboard in edit mode → add the Hexagonal grid widget.
Use: shows the facility's endpoints as hexagons colored by state, with access to details.
Parameters: variable, columns, hexagon size and legend.
Filters: responds to Filter Set chips.

Device widgets by severity

Three new widgets that count devices by their most severe active alarm.

Device counter by severity: one counter per severity, optionally including normal devices.
Device counter (single severity): a single counter for the chosen severity.
Device last state: a table with severity, device, state, date and facility.
The counters include the client's custom severities.

State time summary widget

Use: shows the time accumulated in each state by one or more endpoints over the period.
Format: bars, pie or table, with accumulated time and percentage.
Gaps: periods without data are shown as "no information".

Automations defined in the device model

Rules in the model script: the configuration script can declare automation rules (trigger and steps). They are created automatically when a device of that model is registered, and the description supports {deviceAddress} and {deviceName}.
Reprovisioning: Manager → Devices → select devices → Reprovision automations re-applies the model's rules and alerts without touching endpoints or history. Changed rules are updated, and renamed rules are renamed instead of duplicated.
Optional cleanup: Remove outdated rule versions deletes old rule versions. It never deletes alerts.

Per-device encryption key and cryptography in scripts

For devices that send AES-encrypted data.

Per-device key: if the model declares it, the device form shows an Encryption key field (hexadecimal, 16, 24 or 32 bytes). It is stored encrypted and no API returns it.
Scripts: the crypto object in uplink and downlink scripts provides aesDecrypt and aesEncrypt (CBC or ECB, several paddings) and conversions between hex, bytes, UTF-8 and Base64. The device key is available as device.encryptionKey.

Simplified Chinese

Simplified Chinese (ZH-CN) is now the eighth language in Monitor, Manager and notifications. In multi-language fields it shows up as "中文".

Multi-select and bulk delete in lists

Use: configuration lists add a checkbox per row and an All / None selector, so several items can be deleted at once after a confirmation.
Persistence: the selection is kept while paging, sorting or filtering.
Dashboards: several dashboards can also be deleted at once in Monitor.

Improvements

Device status widget: the donut (formerly "Active alarms") splits devices by their worst active severity plus "Normal", with the dominant percentage in the center.
Alarm counter: new card design with uniform sizing and an option to include "Normal". The "Both" direction was removed; widgets that used it switch to horizontal.
Weather widget: redesigned, with an icon for the current condition, outdoor temperature and the facility's sensors, and better behavior when resized.
Widget groups: new horizontal orientation with adjustable row height.
Font size: Metrics, Rounded Gauge and Device widgets accept a fixed font size; leaving it empty means automatic.
Mobile layout: when the dashboard collapses to one column, it keeps the desktop reading order.
Camera widget: browse images frame by frame with arrows and a time slider.
Simplified table: color ranges are edited in the variable's displayed unit.
Hidden endpoints: dashboards no longer show widgets whose endpoints are all hidden.
Widget picker: new and redesigned thumbnails.
Low severity color: changes from #FFDD00 to #FFDD30 across the platform.
User activity log: moved to the client's Security menu.
White-label favicon: now kept across updates.
Notifications: the email signature is in English for English, French, Italian, German and Chinese (it used to be in Spanish).

API changes

Severities as integers: to support custom severities, severity fields on alerts and alarms (for example Alert.Severity) change from an enum to an integer. Existing severity values are unchanged, but integrations that validate against the fixed list of severities must accept new values.
New services: AIService, AIContextService, AIOnboardingService, AlertTemplateService, DeviceMigrationService, DashboardTemplateService, DashboardPresetService and DashboardFilterSetService.
New methods: alarm lifecycle in AlarmService (GetActiveAlarmsLifecycle, ResolveAlarm, MuteAlarm, ResetAlarm, etc.), custom severities (EnumAlarmTypeSeverities plus create/edit/delete), DashboardService.ExecuteControlWidget, DashboardService.DeleteDashboards, DeviceService.ReprovisionDeviceAutomations, DeviceService.SetDeviceEncryptionKey, and authenticator-app 2FA in AccountService (EnrollTotp, ConfirmTotpEnrollment, DisableTotp).
Rate-of-change alerts: Alert, AlertCondition and Event gain RateOfChangeWindowMinutes, RateOfChangeThreshold and RateOfChangeDirection.

Bug fixes

2FA for administrators only: the "Require 2FA for administrators only" option had no effect. It is now enforced.
Activity log: the scheduled purge of the activity log always failed and deleted nothing.
Alert severity: changing an alert's severity did not update its open alarms.
Simplified table: range colors disappeared on variables with unit conversion, and a deleted endpoint caused the widget to be dropped on save.
Rounded gauge: the half circle was clipped on narrow widgets.
Camera widget: Launch Live did not show up when there were no images yet or in edit mode.
Dashboards: creating an empty client dashboard returned an error.
Unknown routes: an invalid URL in Monitor or Manager left a blank screen; it now redirects to the home page.
Notifications: the escalation push notification was always in English.

Support

For questions, contact the Cloud Studio team: contacto@cloud.studio.

v1.8.4 — June 2026