ICALEPCS2023 - List of Authors (Baker, K.R.L.)

Paper	Title	Page
TU1BCO01	A Workflow for Training and Deploying Machine Learning Models to EPICS	244
	M.F. Leputa, K.R.L. Baker, M. Romanovschi STFC/RAL/ISIS, Chilton, Didcot, Oxon, United Kingdom
	The transition to EPICS as the control system for the ISIS Neutron and Muon Source accelerators is an opportunity to more easily integrate machine learning into operations. But developing high quality machine learning (ML) models is insufficient. Integration into critical operations requires good development practices to ensure stability and reliability during deployment and to allow robust and easy maintenance. For these reasons we implemented a workflow for training and deploying models that utilize off-the-shelf, industry-standard tools such as MLflow. Our experience of how adoption of these tools can make developer’s lives easier during the training phase of a project is discussed. We describe how these tools may be used in an automated deployment pipeline to allow the ML model to interact with our EPICS ecosystem through Python-based IOCs within a containerized environment. This reduces the developer effort required to produce GUIs to interact with the models within the ISIS Main Control Room as tools familiar to operators, such as Phoebus, may be used.
	Slides TU1BCO01 [3.370 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TU1BCO01
About •	Received ※ 05 October 2023 — Accepted ※ 12 October 2023 — Issued ※ 19 October 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP108	Progress of the EPICS Transition at the Isis Accelerators	817
	I.D. Finch, B.R. Aljamal, K.R.L. Baker, R. Brodie, J.-L. Fernández-Hernando, G.D. Howells, M.F. Leputa, S.A. Medley, M. Romanovschi STFC/RAL/ISIS, Chilton, Didcot, Oxon, United Kingdom A. Kurup Imperial College of Science and Technology, Department of Physics, London, United Kingdom
	The ISIS Neutron and Muon Source accelerators have been controlled using Vsystem running on OpenVMS / Itaniums, while beamlines and instruments are controlled using EPICS. We outline the work in migrating accelerator controls to EPICS using the PVAccess protocol with a mixture of conventional EPICS IOCs and custom Python-based IOCs primarily deployed in containers on Linux servers. The challenges in maintaining operations with two control systems running in parallel are discussed, including work in migrating data archives and maintaining their continuity. Semi-automated conversion of the existing Vsystem HMIs to EPICS and the creation of new EPICS control screens required by the Target Station 1 upgrade are reported. The existing organisation of our controls network and the constraints this imposes on remote access via EPICS and the solution implemented are described. The successful deployment of an end-to-end EPICS system to control the post-upgrade Target Station 1 PLCs at ISIS is discussed as a highlight of the migration.
	Poster TUPDP108 [0.510 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP108
About •	Received ※ 02 October 2023 — Accepted ※ 04 December 2023 — Issued ※ 17 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

WE2BCO04	Maintaining a Hybrid Control System at ISIS with a Vsystem/EPICS Bridge	986
	K.R.L. Baker, I.D. Finch, M. Romanovschi STFC/RAL/ISIS, Chilton, Didcot, Oxon, United Kingdom
	The migration of the controls system for the ISIS accelerator from Vsystem to EPICS presents a significant challenge and risk to day-to-day operations. To minimise this impact throughout the transition, a software bridge between the two control systems has been developed that allows the phased porting of HMIs and hardware. The hybrid Vsystem and EPICS system also allows the continued use of existing feedback control applications that now require interaction between both control systems, for example the halo steering operation in Target Station 1. This work describes the implementation of this bridge, referred to as PVEcho, for the mapping of Vsystem channels to EPICS PVs and vice versa. The position within the wider ISIS controls software stack is outlined as well as how it utilises Python libraries for EPICS. Finally, we will discuss the software development practices applied that have allowed the bridge to run reliably for months at a time.
	Slides WE2BCO04 [2.757 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-WE2BCO04
About •	Received ※ 05 October 2023 — Accepted ※ 08 December 2023 — Issued ※ 11 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

Paper

Title

Page

A Workflow for Training and Deploying Machine Learning Models to EPICS

244

M.F. Leputa, K.R.L. Baker, M. Romanovschi
STFC/RAL/ISIS, Chilton, Didcot, Oxon, United Kingdom

The transition to EPICS as the control system for the ISIS Neutron and Muon Source accelerators is an opportunity to more easily integrate machine learning into operations. But developing high quality machine learning (ML) models is insufficient. Integration into critical operations requires good development practices to ensure stability and reliability during deployment and to allow robust and easy maintenance. For these reasons we implemented a workflow for training and deploying models that utilize off-the-shelf, industry-standard tools such as MLflow. Our experience of how adoption of these tools can make developer’s lives easier during the training phase of a project is discussed. We describe how these tools may be used in an automated deployment pipeline to allow the ML model to interact with our EPICS ecosystem through Python-based IOCs within a containerized environment. This reduces the developer effort required to produce GUIs to interact with the models within the ISIS Main Control Room as tools familiar to operators, such as Phoebus, may be used.

Slides TU1BCO01 [3.370 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TU1BCO01

About •

Received ※ 05 October 2023 — Accepted ※ 12 October 2023 — Issued ※ 19 October 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP108

Progress of the EPICS Transition at the Isis Accelerators

817

I.D. Finch, B.R. Aljamal, K.R.L. Baker, R. Brodie, J.-L. Fernández-Hernando, G.D. Howells, M.F. Leputa, S.A. Medley, M. Romanovschi
STFC/RAL/ISIS, Chilton, Didcot, Oxon, United Kingdom
A. Kurup
Imperial College of Science and Technology, Department of Physics, London, United Kingdom

The ISIS Neutron and Muon Source accelerators have been controlled using Vsystem running on OpenVMS / Itaniums, while beamlines and instruments are controlled using EPICS. We outline the work in migrating accelerator controls to EPICS using the PVAccess protocol with a mixture of conventional EPICS IOCs and custom Python-based IOCs primarily deployed in containers on Linux servers. The challenges in maintaining operations with two control systems running in parallel are discussed, including work in migrating data archives and maintaining their continuity. Semi-automated conversion of the existing Vsystem HMIs to EPICS and the creation of new EPICS control screens required by the Target Station 1 upgrade are reported. The existing organisation of our controls network and the constraints this imposes on remote access via EPICS and the solution implemented are described. The successful deployment of an end-to-end EPICS system to control the post-upgrade Target Station 1 PLCs at ISIS is discussed as a highlight of the migration.

Poster TUPDP108 [0.510 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP108

About •

Received ※ 02 October 2023 — Accepted ※ 04 December 2023 — Issued ※ 17 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

WE2BCO04

Maintaining a Hybrid Control System at ISIS with a Vsystem/EPICS Bridge

986

K.R.L. Baker, I.D. Finch, M. Romanovschi
STFC/RAL/ISIS, Chilton, Didcot, Oxon, United Kingdom

The migration of the controls system for the ISIS accelerator from Vsystem to EPICS presents a significant challenge and risk to day-to-day operations. To minimise this impact throughout the transition, a software bridge between the two control systems has been developed that allows the phased porting of HMIs and hardware. The hybrid Vsystem and EPICS system also allows the continued use of existing feedback control applications that now require interaction between both control systems, for example the halo steering operation in Target Station 1. This work describes the implementation of this bridge, referred to as PVEcho, for the mapping of Vsystem channels to EPICS PVs and vice versa. The position within the wider ISIS controls software stack is outlined as well as how it utilises Python libraries for EPICS. Finally, we will discuss the software development practices applied that have allowed the bridge to run reliably for months at a time.

Slides WE2BCO04 [2.757 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-WE2BCO04

About •

Received ※ 05 October 2023 — Accepted ※ 08 December 2023 — Issued ※ 11 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)