Who ThingsBoard is for#
Industrial IoT engineers building SCADA or monitoring systems
ThingsBoard ships SCADA symbol libraries and dashboard templates for process visualization. Engineers can model a plant's valves, pumps, and tanks as digital twins, build real-time control screens, and set alarm rules that trigger on any signal deviation. The rule engine connects alarms to external notification systems without custom integration code.
Skip if:
If your use case requires only device connectivity and a time-series store with no dashboard or alarm requirement, a lighter stack (MQTT broker plus InfluxDB) may be simpler to operate.
IoT product companies shipping white-labeled customer dashboards
ThingsBoard supports multi-tenancy: each customer gets their own device namespace, permissions, and dashboard view. Dashboards are white-labeled at the tenant level. Product companies can ship branded IoT experiences to their customers without building a separate frontend or buying a dashboard tier from a cloud provider.
Skip if:
If you need only a single-tenant deployment for internal use and white-labeling is not a requirement, the multi-tenancy overhead adds setup complexity that a simpler single-tenant platform avoids.
Smart energy and utilities teams managing meters at scale
ThingsBoard's smart metering path covers collecting data from meters, storing it reliably, aggregating consumption statistics, detecting anomalies and fraud, and presenting results to administrators and end users. SCADA and data visualization templates are available for energy distribution and generation monitoring.
Skip if:
If your metering deployment uses a proprietary protocol not in ThingsBoard's connectivity stack and you cannot deploy a gateway layer, a custom integration path may be needed before the platform is useful.
The problem it solves#
Building IoT infrastructure means solving device connectivity, protocol normalization, time-series storage, alarm processing, and dashboard delivery as separate engineering problems. Commercial IoT platforms bundle these layers, but the cost structure creates its own challenge: Azure IoT Central and AWS IoT charge per device and per message, so a deployment of tens of thousands of sensors generates bills that scale with fleet growth rather than flattening as infrastructure costs normally do.
The deeper issue is data ownership. When telemetry flows into a managed cloud service, the raw data lives on the vendor's infrastructure. Exporting historical data at scale, building custom processing pipelines, or integrating with on-premises business systems requires working around restrictions the vendor controls. Teams that need white-labeled dashboards for their own customers face additional licensing tiers on paid platforms.
How it solves it#
Multi-protocol device connectivity
Connects devices over MQTT, CoAP, HTTP, LwM2M, Modbus, OPC UA, LoRaWAN, BACnet, Sigfox, SNMP, and S7 natively. Gateways and platform integrations cover additional industrial protocols. Devices, sensors, and machines from different vendors can coexist in one deployment without a custom translation layer.
Rule chains and alarm engine
Rule chains process incoming telemetry in real time: filter signals, transform payloads, trigger alarms on threshold breaches or device inactivity, and push notifications or data into external systems including CRM, ERP, email, and SMS. Alarm rules and notification rules are configured from the UI without writing code.
Real-time dashboard builder with 600+ widgets
Dashboards are assembled from over 600 widgets covering charts, gauges, maps, tables, SCADA symbols, and device controls. Operators can view live telemetry and send commands from the same screen. Dashboards can be white-labeled and shared directly with end customers under their own branding.
Digital twin asset modeling
Devices, assets, and customers are organized into a hierarchy with defined relations between them. This models real-world structures (warehouse, zone, unit, sensor) so that permissions, data aggregation, and alarms operate at the correct level, whether a single sensor or an entire site.
Monolith to cluster scaling
A single ThingsBoard instance handles up to approximately one million devices on PostgreSQL or Cassandra. Adding nodes for a clustered deployment uses Kafka and Valkey; device profiles, dashboards, and alarm rules carry over unchanged. Performance scales linearly as nodes are added.
Strengths and trade-offs#
Strengths
- Full stack from device connection to customer dashboardThingsBoard covers connectivity, data modeling, processing, alarm rules, and dashboard delivery in one deployment. Most open source IoT tools solve one layer (a broker, a time-series database, or a dashboarding tool) and require integration work across the stack. ThingsBoard ships the entire path.
- Broad protocol support without a separate gateway layerNative support for MQTT, CoAP, HTTP, LoRaWAN, Modbus, OPC UA, BACnet, LwM2M, S7, Sigfox, and SNMP means most industrial and consumer device types connect without a custom gateway. That coverage is built into the platform, not a paid add-on.
- Actively maintained with 22,500+ GitHub starsThe repository had its last push on 2026-10-01 and has accumulated 22,500+ stars and 6,400+ forks since its 2016 launch. Active development, a documented versioning cadence (BUSL releases convert to Apache 2.0 after four years), and a large contributor base reduce the abandonment risk common with smaller IoT projects.
Trade-offs
- -BUSL 1.1 prohibits reselling as a hosted serviceVersion 4.4 and later ship under the Business Source License 1.1. You can run ThingsBoard on your own infrastructure for internal production use at no cost within the Additional Use Grant. What is not permitted: offering ThingsBoard as a managed or hosted IoT service to third parties, or removing ThingsBoard branding, without a commercial license. Each BUSL release converts to Apache 2.0 four years after it ships. Versions before 4.4 remain Apache 2.0 with no such restrictions.
- -Java-based, heavier resource footprint than lightweight brokersThingsBoard is a Java microservices platform. The monolith requires more memory and CPU than a lightweight MQTT broker or a single-purpose time-series database. Cluster mode adds Kafka, Valkey, and optionally Cassandra, which raises infrastructure complexity and resource requirements further. Teams looking for a minimal-footprint IoT broker should evaluate lighter alternatives.
ThingsBoard vs alternatives#
ThingsBoard vs Azure IoT Central
Azure IoT Central is Microsoft's managed IoT application platform. Both handle device management, telemetry collection, and dashboards, but the deployment model differs entirely. Azure IoT Central is a SaaS product with no self-hosting option; ThingsBoard runs on your own infrastructure.
| Feature | ThingsBoard | Azure IoT Central |
|---|---|---|
| License | BUSL 1.1 (v4.4+); Apache 2.0 (earlier) | Proprietary |
| Self-hosting | Yes | No |
| Pricing model | Free self-hosted within Additional Use Grant | Per-device/per-message fees |
| Protocol support | MQTT, CoAP, HTTP, LoRaWAN, Modbus, OPC UA, and more | MQTT, AMQP, HTTPS |
| White-label dashboards | Yes | Limited |
ThingsBoard is the stronger choice when you need data sovereignty, want to avoid per-device fees at scale, or require industrial protocols like Modbus, OPC UA, or BACnet that Azure IoT Central does not natively support. Azure IoT Central is worth choosing when your team lacks infrastructure capacity and you need Azure ecosystem integrations (Power BI, Event Hub, Logic Apps) out of the box with no server management.
ThingsBoard vs AWS IoT
AWS IoT is a suite of managed services (IoT Core, IoT Greengrass, IoT SiteWise) rather than a single platform. ThingsBoard covers roughly the same functional scope in one self-hosted deployment.
| Feature | ThingsBoard | AWS IoT |
|---|---|---|
| License | BUSL 1.1 (v4.4+) | Proprietary |
| Self-hosting | Yes | No |
| Dashboard layer | Built-in (600+ widgets) | Requires Grafana or QuickSight add-on |
| Industrial protocol support | Native (Modbus, OPC UA, BACnet, etc.) | Via Greengrass (additional setup) |
| Multi-tenancy | Built-in | Requires custom IAM design |
ThingsBoard avoids the integration overhead of stitching AWS services together and includes a full dashboard and white-label layer by default. AWS IoT is worth choosing when your workloads already live in AWS, you need managed reliability at very large scale, or you need Lambda and other AWS services tightly integrated into your processing pipeline.
What it's built on#
- Languages
- JavaJavaScriptTypeScript
- Frameworks
- AngularExpress
- Messaging
- KafkaMQTT
FAQ#
Is ThingsBoard free to self-host?
Yes, with conditions. ThingsBoard 4.4 and later ship under the Business Source License 1.1, which permits self-hosting for internal production use at no cost within the Additional Use Grant. Reselling it as a managed service or removing ThingsBoard branding requires a commercial license. Versions before 4.4 remain Apache 2.0 with no such restrictions. ThingsBoard Cloud is the hosted option for teams that prefer not to manage infrastructure.
How does ThingsBoard compare to Azure IoT Central and AWS IoT?
Azure IoT Central and AWS IoT are managed cloud services: Microsoft and Amazon handle infrastructure, but telemetry is stored in their systems and pricing scales with device count and message volume. ThingsBoard runs on your own infrastructure, so you control where data lives and there are no per-device fees on the self-hosted path. The tradeoff is that you manage the server or cluster. ThingsBoard is the better fit when data sovereignty, cost at scale, or protocol flexibility is the priority; Azure and AWS are worth choosing when managed reliability and tight ecosystem integrations matter more.
What connectivity protocols does ThingsBoard support?
ThingsBoard connects devices over MQTT, CoAP, HTTP, LwM2M, Modbus, OPC UA, LoRaWAN, BACnet, Sigfox, SNMP, and S7 natively. IoT gateways and platform integrations extend this to additional industrial and consumer device types. Most sensor and industrial equipment categories have a supported connectivity path without requiring a custom gateway.
Can I white-label ThingsBoard dashboards for my customers?
Yes. ThingsBoard supports multi-tenancy with per-customer branding. Each customer tenant gets their own device namespace, permissions scope, and dashboard configuration. Product companies use this to ship branded IoT monitoring experiences to their end customers from a shared ThingsBoard deployment. White-labeling for customer-facing use is supported in the self-hosted version.
Similar open-source tools#
RuView
Intelligent AI agents for real-world applications
FckSignups
Open-source tools that work instantly, no signup required
Paymenter
Self-hosted billing for hosting providers, MIT licensed
3x Ui
Web panel for Xray-core with per-client and multi-node controls
Vela
OpenStreetMap and Google data, no account, no Play Services
Esp32 C3 Adblock
Pi-hole-style DNS ad-blocker on a $2 ESP32-C3 board