Milestone AI Bridge containers

Milestone AI Bridge consists of multiple container images with each image used for specific functions within the entire solution. The container images are a part of the deployment of Milestone AI Bridge but are also available for individual download.

To facilitate deployment of all these containers, you can use the Milestone AI Bridge Helm chart to install the containers on multiple processing servers using Kubernetes or install them on a single processing server using Docker Compose by running the docker-compose-production.yml file.

If necessary, the default settings of the .yaml file inside the Milestone AI Bridge Helm chart or the .yaml file for deployment on a single server using Docker Compose can be edited, enabling you to fine-tune the deployment of Milestone AI Bridge on your processing server.

Milestone AI Bridge has an event driven architecture and much of the communication between the containers takes place through brokers and topics provided by Apache Kafka.

Name	Role
AI Bridge Streaming (aibridge-streaming)	This container grants IVA applications access to video streams from the XProtect VMS using the RTSP protocol and a gRPC based protocol named Direct Streaming. Direct Streaming enables access to live and recorded video. Use the aibridge-webservice container to query which protocol to use as well as the protocol specific endpoints.
AI Bridge Kafka zookeeper (aibridge-kafka-zookeeper)	This container runs a Kafka Zookeeper instance that keeps track of all the brokers and Apache Kafka topics. The Docker image for this container is also available here: https://hub.docker.com/r/confluentinc/cp-zookeeper/.
AI Bridge Kafka broker (aibridge-kafka-broker)	This container runs a Kafka Broker instance hosting the Apache Kafka topics. The Docker image for this container is also available here: https://hub.docker.com/r/confluentinc/cp-kafka/.
AI Bridge Fuseki (aibridge-fuseki)	This container runs the Apache Jena Fuseki SPARQL server using TBD for a RDF storage database. The database replicates parts of the XProtect VMS configuration so the GraphQL interface exposed by the aibridge-webservice container can be queried in a database. The database only functions as a cache and no data needs to be persisted. The database is only populated when Milestone AI Bridge is initialized. Changes made to the XProtect VMS configuration, for example if a camera name has been changed or the Field of View has been enabled or disabled, will be reflected in the database as well.
AI Bridge Init (aibridge-init)	This container initializes the Milestone AI Bridge and registers it as a service in XProtect VMS. After the registration, the processing servers and the IVA applications are available for configuration and subscription in XProtect Management Client and the container will be stopped.
AI Bridge Connector (aibridge-connector)	This container connects to and receives data from the XProtect VMS. The type of data will determine the protocol used to retrieve the data. The container handles sharing of the configuration data, status of devices, and video streams from the XProtect VMS to Milestone AI Bridge.
AI Bridge Webservice (aibridge-webservice)	This container is the main entry point for the IVA application and Milestone AI Bridge interaction. This container exposes the GraphQL interface so camera details in the XProtect VMS can be queried by the IVA application. The container enables the IVA application to: query the availability of protocol specific endpoints (e.g. RTSP) for getting video and other data from the XProtect VMS. query the availability of event, metadata, and video topics that the IVA application can use to send generated data back into the XProtect VMS.
AI Bridge Health (aibridge-health )	This container exposes an API that enables the administrator of XProtect Management Client to monitor the health of the processing servers and the connectivity between the XProtect VMS and Milestone AI Bridge.
AI Bridge Broker (aibridge-broker)	This container acts as a broker to which an IVA application can send events, metadata, and video. The data is sent to specific topics, which another container (aibridge-proxy) will subscribe to in order to forward data back into the XProtect VMS. The broker supports both a REST and gRPC API for sending data to the topics.
AI Bridge Proxy (aibridge-proxy)	This container sends events, metadata, and video back into the XProtect VMS. In the XProtect VMS, events can be used by the built-in rule engine to trigger actions. You can trigger recording of videos when an event occurs. You can also trigger an alarm which an operator then can manage through the built-in Alarm Manager. Frame-based metadata can also be sent back into the XProtect VMS and stored. For example, a detected object can be highlighted on the video, not only for live video, but also for recorded video. The built-in search functionality can also use the metadata to locate relevant videos using different search criteria.

Name

Role

AI Bridge Streaming

(aibridge-streaming)

This container grants IVA applications access to video streams from the XProtect VMS using the RTSP protocol and a gRPC based protocol named Direct Streaming. Direct Streaming enables access to live and recorded video.

Use the aibridge-webservice container to query which protocol to use as well as the protocol specific endpoints.

AI Bridge Kafka zookeeper

(aibridge-kafka-zookeeper)

This container runs a Kafka Zookeeper instance that keeps track of all the brokers and Apache Kafka topics.

The Docker image for this container is also available here: https://hub.docker.com/r/confluentinc/cp-zookeeper/.

AI Bridge Kafka broker

(aibridge-kafka-broker)

This container runs a Kafka Broker instance hosting the Apache Kafka topics.

The Docker image for this container is also available here: https://hub.docker.com/r/confluentinc/cp-kafka/.

AI Bridge Fuseki

(aibridge-fuseki)

This container runs the Apache Jena Fuseki SPARQL server using TBD for a RDF storage database.

The database replicates parts of the XProtect VMS configuration so the GraphQL interface exposed by the aibridge-webservice container can be queried in a database.

The database only functions as a cache and no data needs to be persisted. The database is only populated when Milestone AI Bridge is initialized.

Changes made to the XProtect VMS configuration, for example if a camera name has been changed or the Field of View has been enabled or disabled, will be reflected in the database as well.

AI Bridge Init

(aibridge-init)

This container initializes the Milestone AI Bridge and registers it as a service in XProtect VMS.

After the registration, the processing servers and the IVA applications are available for configuration and subscription in XProtect Management Client and the container will be stopped.

AI Bridge Connector

(aibridge-connector)

This container connects to and receives data from the XProtect VMS.

The type of data will determine the protocol used to retrieve the data. The container handles sharing of the configuration data, status of devices, and video streams from the XProtect VMS to Milestone AI Bridge.

AI Bridge Webservice (aibridge-webservice)

This container is the main entry point for the IVA application and Milestone AI Bridge interaction.

This container exposes the GraphQL interface so camera details in the XProtect VMS can be queried by the IVA application.

The container enables the IVA application to:

query the availability of protocol specific endpoints (e.g. RTSP) for getting video and other data from the XProtect VMS.
query the availability of event, metadata, and video topics that the IVA application can use to send generated data back into the XProtect VMS.

AI Bridge Health

(aibridge-health )

This container exposes an API that enables the administrator of XProtect Management Client to monitor the health of the processing servers and the connectivity between the XProtect VMS and Milestone AI Bridge.

AI Bridge Broker

(aibridge-broker)

This container acts as a broker to which an IVA application can send events, metadata, and video.

The data is sent to specific topics, which another container (aibridge-proxy) will subscribe to in order to forward data back into the XProtect VMS.

The broker supports both a REST and gRPC API for sending data to the topics.

AI Bridge Proxy

(aibridge-proxy)

This container sends events, metadata, and video back into the XProtect VMS.

In the XProtect VMS, events can be used by the built-in rule engine to trigger actions.

You can trigger recording of videos when an event occurs. You can also trigger an alarm which an operator then can manage through the built-in Alarm Manager.

Frame-based metadata can also be sent back into the XProtect VMS and stored. For example, a detected object can be highlighted on the video, not only for live video, but also for recorded video.

The built-in search functionality can also use the metadata to locate relevant videos using different search criteria.

For more information about where to download Milestone AI Bridge resources, see Milestone AI Bridge resource location.