Background

The is an international neutrino experiment aiming to increase our understanding of the fundamental building blocks of the universe. The setup will consist of a powerful neutrino emitter aimed through two underground detector installations. One installation will be located near the surface at the Fermi National Accelerator Laboratory in the USA, and another larger detector will be located over a kilometre underground at the Sanford Underground Research Facility.

An important component of the experiment is the DUNE Run Control (DRunC) software, which orchestrates the many instruments that monitor and record the vast amounts of data the experiment is expected to generate. In addition of the main finite state machine (FSM) interface used to control the experiment, a lower level 'expert' interface allows engineers finer-grained control to components for troubleshooting and testing.

Our Contribution

In this first collaboration with the DUNE team, we developed a Django-based UI which exposes with the high-level FSM and low-level 'expert' interface as a deployable Python web server. Best practice workflows were adopted to ensure code is modern, clean and testable throughout, and able to be deployed inside a standalone Docker container instance.

Using a Postgres database backend, we implemented permissions-based access to the two interfaces, as well as the collection and caching of operational state and messages from instrumentation server infrastructure. Using this state, we constructed an interactive diagram-based visualisation and control of the main FSM, and created an expert-mode live remote process manager and message log view interface for applications running in the instrumentation servers.

Screenshots

DUNE finite state machine

The Controller, showing the Finite State Machine diagram

DUNE process control

The Process Manager

DUNE control webpage

The Session Manager

Outcomes

A feature-complete prototype of the web frontend was deployed and presented to the DUNE team. The project was transferred over and incorporated into the , together with documentation and user support for its operation. We have since secured further funding for ImperiaRSE development and consultation throughout the wider DUNE software stack, continuing to this day, with further funding expected in the near future.