Connector bees
Bridge BeeOS with the outside world — translating protocols and bringing data in and out of the platform
Connector bee agents are responsible for integrating BeeOS with external systems and services. They communicate using specific protocols and translate incoming or outgoing messages into a unified format that can be processed by the hive and other bee agents. These agents enable seamless interoperability between industrial devices, cloud platforms, and enterprise applications.
Connector bees can interface with a wide range of services, including databases, IoT platforms, messaging systems, and cloud data lakes. Different connector bees support different protocols — such as HTTP, MQTT, OPC-UA, or other industrial and web standards — allowing the platform to adapt to diverse environments without requiring custom integrations for every use case.
The platform delivers a powerful catalog of agents that connect systems, process data, and run distributed workloads, while the open Bee Core SDK allows developers to create and extend new agents without limits.
Open by design. Built to connect everything — designed to be extended by anyone
Connectors
A bee agent is comprised of connectors, each of which operates independently. The connectors receive their own configuration from the hive and carry out their respective functions without knowledge of the other connectors. This allows for a single bee agent to connect to multiple sources.
There are bee agents that offer various types of connectors, such as the HTTP Bee agent, where some connectors can function as HTTP clients and others as HTTP servers.
Channels
Each bee agent is composed of connectors and each connector is made up of channels. In BeeOS, channels act as signal interfaces that allow bee agents to exchange data through a unified messaging model. Rather than communicating directly, agents emit and receive signals through channels, enabling loosely coupled interactions across distributed environments.
Each channel represents a defined group of signals associated with a topic, which the hive node uses to route messages between agents.
This approach abstracts the complexity of the underlying messaging system, allowing developers to focus on defining how agents produce or consume signals rather than how communication is implemented internally.
Channel types
Depending on their role in the signal flow, channels are divided into two primary types:
Emitter channels
Emitter channels are used by bee agents to publish signals into the platform. When an agent generates new information — such as processed data, workload results, or automation events — it emits those signals through an emitter channel. The hive node routes these emitted signals to any receiver channels subscribed to the corresponding topic, enabling asynchronous and scalable data flows.
Receiver channels
Receiver channels allow bee agents to listen for signals produced by other agents. By subscribing to a topic, an agent can react to incoming data, trigger logic, or continue a processing pipeline without requiring direct connections to the emitting agent.
Signals
A signal represents the smallest unit of data transmitted within the platform. It contains a single value enriched with metadata such as timestamp, value type, or contextual information. Signals provide a standardized representation for measurements, computed values, or events generated by bee agents.
Signals are grouped together into messages, which are published to a channel topic and routed by the hive node to other agents. By separating signals from the messaging layer, BeeOS ensures that data remains structured and consistent regardless of how it is transported.
Channel Mapping
Channels include a mapping layer that translates signals into the specific format required by the protocol implemented by a bee agent. This mapping allows the platform to maintain a consistent internal signal model while adapting communication to external systems or specialized workloads.
When an agent emits data through an emitter channel, the mapping process transforms the collected information into a standardized signal message and publishes it to the corresponding topic. On the receiving side, the receiver agent applies its mapping configuration to convert incoming signals into the protocol-specific data format required by the external service connected to the agent.
This abstraction ensures that bee agents can exchange information using a unified signal interface, regardless of whether the underlying communication relies on HTTP, MQTT, industrial protocols, or custom integrations. By separating signal representation from protocol formatting, BeeOS enables flexible interoperability while keeping agent logic simple and reusable.
Last updated