Doberman: a modular and distributed slow control system for small- to medium-scale experiments

TL;DR

Doberman addresses the need for a lightweight yet robust slow control solution for small- to medium-scale experiments by delivering a modular, open-source architecture that supports heterogeneous devices, distributed deployment, automation, and alarm handling. The system combines a dual-database backend (MongoDB for configuration and logs, InfluxDB for time-series data), a multi-threaded Monitor framework coordinated by a Hypervisor, and a plugin-based device layer with a pipeline-based automation layer. A web-based interface, Doberview, enables live monitoring, configuration, and controlled interaction, including an interactive overview, detailed sensor views, pipeline management, and embedded visualization tools. Validations across GeMSE, XeBRA, and PANCAKE demonstrate scalability from single-host operation to large, multi-instrument facilities, with open-source licenses and a growing ecosystem of device plugins and documentation for broad adoption.

Abstract

We present Doberman (Detector OBsERving and Monitoring ApplicatioN), a lightweight, modular, and open-source slow control system designed for small-to medium-scale physics experiments. Doberman addresses the gap between heavyweight industrial SCADA frameworks and ad hoc laboratory solutions by providing a flexible software architecture that supports heterogeneous instrumentation, distributed deployment, automated control, and robust alarm handling. The web-based graphical user interface Doberview provides live and continuously updated visualization, configuration, and control of the entire experiment, supporting both routine operation and rapid response to exceptional conditions. Doberman has been deployed and validated in multiple experimental setups, ranging from a remotely operated underground gamma-ray spectrometer to a large, highly instrumented liquid xenon test facility with several hundred monitored quantities. Doberman and Doberview are released under permissive open-source licenses, and the software, documentation, and example device integrations are publicly available.

Figures (8)

• Figure 1: Overview of a distributed Doberman setup. A main server hosts the databases and, optionally, a network-mounted file system. The Hypervisor (central coordination) and PipelineMonitors (automation) typically run on the main server, while DeviceMonitors (interface to measuring devices) can be distributed across additional hosts to reduce cable lengths and improve resilience.
• Figure 2: Overview of communication between Doberman Monitors and the storage database. Orange arrows indicate data flow and pink arrows command flow. Each box inside a Monitor represents a thread. All Monitors additionally access the configuration database (not shown).
• Figure 3: Examples of Doberman pipelines: a simple threshold alarm (top), a conversion pipeline computing a derived quantity (center), and a control pipeline actuating a valve based on sensor inputs (bottom).
• Figure 4: Interactive overview in Doberview for the PANCAKE experiment Brown:2023vgf. Sensor boxes display current values, pumps and valves change color or animate according to their state, and liquid levels are visualized based on derived quantities.
• Figure 5: Example of a sensor detail window in Doberview. Configuration parameters and alarm settings are shown on the left, while recent measurement history and associated pipelines are displayed on the right.