Wayne Salter, CERN IT-CO JCOP Technology Survey (SCADA) JCOP External Review 10 th -12 th March 2003.

Wayne Salter, CERN IT-CO

JCOP Technology Survey(SCADA)

JCOP External Review10th-12th March 2003

22W. Salter JCOP External Review 10th-12th March 2003

Overview of Presentation

BackgroundProcessSelection criteriaResult of tenderFeatures of chosen productRequired enhancementsConclusions


Background Looking for suitable common controls

infrastructure for LHC experimentsReduced resources for controlsTo achieve a homogeneous control system

EPICS/TIS 4000 evaluations (1996/1997) were not entirely positive

Componentware did not look mature enough

Some initial experience with SCADA systems

1997 began to look at commercial market (Technology Survey)Led by Axel Daneels with 10 other staff from IT-CO

and one coopérant from ALICECan SCADA systems be used or not?


What was the process followed?


Process - I Identified possible suppliers (~150) Issued a request for information

(questionnaire) Based on the responses (~40) produced a

short list (~20) Visited short listed companies with a

detailed questionnaire Ranked products based on criteria and

weightings agreed within the JCOP EB Selected 5 for hands-on evaluation,

visited reference sites and discussed evolution of products with the vendors


Process - II Performed detailed evaluation of

selected products Continued to watch the market and

as a result select a further product for evaluation

Presented results at JCOP Workshop II (September 1999)

Obtained agreement from the LHC experiments to issue a tender

Market survey launched in December 1999


What were the selection

criteria used?


Controls Technologies (LHC Era )

Supervision

ProcessManagement

FieldManagement

Sensors/devices

Field buses & Nodes

PLC/UNICOS

OPC

Communication Protocols

SCADA

Technologies

Experimental equipment

LAN

WAN

Storage

Oth

er

syst

em

s(L

HC

, S

afe

ty,

...)

Configuration DB,Archives,Log files, etc.

Controller/PLC VME

Field Bus

LAN

Node Node

VME

DIM

FSM

Commercial Custom

Based on an original idea from LHCb

Layer Structure


SCADA Selection for the LHC Experiments (I)

ArchitectureEvent drivenDevice oriented

Scalability Static

1 Mio Input/Output channels

Large no. of alarms and archive data

Large no. of stations Dynamic

Ability to add to the system No limitations

No. of stationsNaming...

Possibility to distribute system over many CPUs

168 Criteria in a number of categories Concentrated only on non-standard features

Openness Access to all internal data

via an Application Programming Interface (API)

Database access Hardware access

OPCCERN Selected PLCsCERN Selected Field Buses


SCADA Selection for the LHC Experiments (II)

FlexibilityCustomisationEasy integration of

user functionality

Development Multi-user teams Distributed teams

CERN supported operating systems WNT, W2000 Linux RedHat 6.1 Inter-operability

Functionality Scripting Array handling Advanced HMI (dynamic) Sophisticated alarm handling Administration of the SCADA

Configuration capabilities Large number of devices Not SCADA specialists Compatibility with other

systems

Operational concept Integrated Stand-alone partitioning Central and remote Non-professional operators Multiple simultaneous

distributed operators


What was the result

of the tender?


Selected SCADA System PVSS II (Prozeßvisualisierungs- und

Steuerungs-System) from ETM Contract negotiated with ETM and

signed in November 2000Unlimited licenses for the LHC experiments

both at CERN and external institutesLicense cost shared between LHC

experiments and IT-COMaintenance paid by IT-CO

In June 2002 extended to CERN-wide agreement


What are the key

features of PVSS?


SCADA - Features of PVSS

Architecture Device structuring Completely event driven Internal data stored in Data

Points PVSS ‘tools’ are standard PVSS

panels with associated scripts

Openness Windows and Linux API / Driver development kit OPC, DDE, ActiveX ADO I/F (ODBC and OLE DB

compliant DBs) Http server with templates Information server ASCII manager ASCII graphics format XML Inclusion of compiled function or

launch an application from a script

Scalability “Unlimited” number of variables Scattered PVSS systems Distributed PVSS systems (up to

127)

Flexibility Everything modifiable on-line Control of PVSS tools via

internal variables Powerful scripting (~Ansi C) Dynamic arrays and sets

Configuration Possibility to use PVSS to

configure itself


PVSS S/W Architecture

User Interface LayerUser Interface Layer

ProcessingProcessing LayerLayer

DriverDriver LayerLayer

CtrlCtrl APIAPI

EVEV

DD DDDD

UIMUIM UIMUIM UIMUIM

DMDM Communication and Communication and Memory LayerMemory Layer


Data Point Concept

Data Point Type

Data Point


DpName

ParentReference

BoardType

$sDpName

$sParentReference

$sBoardType

ReferencePanel

Graphics’ Concept

dummy:myFirstCaenCrateSy403S3

dummy:myFirstCaenCrateSy403

A504N

dummy:myFirstCaenCrateSy403S1

dummy:myFirstCaenCrateSy403

A503P


Scattered System

TCP/IP

NT/2000

UIM

NT/2000

UIM

Linux

UIM

Linux

Ctrl

Linux

API

NT/2000

Driver

Linux

Driver

Linux

EV+DM


Distributed Systems

CTRLCTRL APIAPI

EVEV

DD DDDD

UIMUIM UIMUIM

DMDM

CTRLCTRL APIAPI

EVEV

DD DDDD


DMDM

CTRLCTRL APIAPI

EVEV

DD DDDD


DMDM DISTDIST DISTDIST

DISTDIST

System1

System3

System2


Redundancy


What needs to be

improved?


Needed Enhancements• Like all SW, PVSS II is also not perfect

Improvements have been identified and discussed in SASG/FW WG/PVSS Users’ Mtg.Some are HEP specific

Regular discussions held with ETMKey items still under discussion include:

DPT synchronisation (PVSS II v3.0)Improved archiving and retrievalRemote access and securitySome aspects of performance (PVSS II v3.0)Trending (some in PVSS II v3.0)Documentation (on-going)

Many requested enhancements already implemented or foreseen in future releases

“Good and close collaboration with ETM”


Conclusion This was a very major effort (~12 Man-

Years) over more than three yearsNeeded to get up to speed with the technologyMostly short-term staff

All experiments chose to use a common commercial solution

PVSS II from ETM chosenUnlimited licenses for LHC experimentsExcellent relationship with companyMany enhancements incorporated into the productSome additional enhancements still required

Provides a good basis for the on-going FW development

Wayne Salter, CERN IT-CO JCOP Technology Survey (SCADA) JCOP External Review 10 th -12 th March 2003.

Documents