Introducing python into industrial environment applications Fabio Pliger SIA s.r.l.

Industrial Supervision and Scada Frameworks process visualization, data acquisition and supervisory control process visualization, data acquisition and supervisory control allow to control and monitor productive units around allow to control and monitor productive units around Client/Sever distributed and open architecture Client/Sever distributed and open architecture Real Time Information Portal Real Time Information Portal

Industrial Supervision and Scada Frameworks(2) Collect “quality” data (both analogic and digital) Collect “quality” data (both analogic and digital) Alarms/Events management Alarms/Events management consistent performance and known response times consistent performance and known response times audit trails in order to maintain a tight control over their operation audit trails in order to maintain a tight control over their operation security and accountability security and accountability

BRIEF ARCHITETURE SCHEME (really really semplified..for iFix) PLCs/Macchines Data Table (DIT) Historical Data Collection Real time data monitoring Events/Alarms management SAC … Engine

We need more… Scada frameworks are great but people need more and more information. So, we need new applications that can perform those tasks. Scada frameworks are great but people need more and more information. So, we need new applications that can perform those tasks. + features = + $$$$$$ + features = + $$$$$$ Our cusotmers loves to have personal solutions! They feel they are important.. Our cusotmers loves to have personal solutions! They feel they are important..

… is python what we need? Before 3 years ago applications were always C and VB… Before 3 years ago applications were always C and VB… 3 years ago: can we do better? 3 years ago: can we do better? Hey, this python language.. Looks realy cool! Hey, this python language.. Looks realy cool!

… Why python? It’s cross platform (???) It’s cross platform (???) time reduction developing new projects time reduction developing new projects Debug is easy Debug is easy Dynamic typing Dynamic typing …but we need 2 different aproaches: …but we need 2 different aproaches: 1. Let’s use it now(!) for new applications with a deadline already fixed 2. Oh.. I think it’s better if we stress and test python before we use in our “market” applications

Aproach 1 We decided this aproach when we had particular situations like: No high performance needs No high performance needs Time is running low… fast development is needed! Time is running low… fast development is needed! Other languages didn’t satisfy us… Other languages didn’t satisfy us…

Aproach 1 - Results We used python to develope some data plotting applications. We used python to develope some data plotting applications. Chaco il great! Chaco il great! It was “easy” to wrap SCADA dlls and access/manage to its data from python It was “easy” to wrap SCADA dlls and access/manage to its data from python It was easy extend, manipulate and add information to scada data. It was easy extend, manipulate and add information to scada data.

Aproach 1 – Results(2) We had a new powerful application We had a new powerful application customer was so satisfied with that application that suddenly asked us to add new features customer was so satisfied with that application that suddenly asked us to add new features Ok! We are ready to try it for other apps! Ok! We are ready to try it for other apps!

Aproach 1 – The Batch Report production batch cycle can be really complex production batch cycle can be really complex it is important to track these production processes (talking about pharmaceutical companies it’s a “must”) it is important to track these production processes (talking about pharmaceutical companies it’s a “must”) Scada frameworks are useful but poor for batch tracking… Scada frameworks are useful but poor for batch tracking…

Aproach 1 – The Batch Report Generator we did is a high classification of the structure of all the production system and its actors, variables, subsystems… we did is a high classification of the structure of all the production system and its actors, variables, subsystems… extra data is stored in a database outside the scada database and defines all the connections between machines, tanks, sensors, variables… extra data is stored in a database outside the scada database and defines all the connections between machines, tanks, sensors, variables… map objects in the production process with their respective variables in the scada database map objects in the production process with their respective variables in the scada database program can run outside the scada system, monitor what is happening and collect data program can run outside the scada system, monitor what is happening and collect data digitals and alarms are event based and analog variables are sampled at a predefined frequency >> time strategies are required! digitals and alarms are event based and analog variables are sampled at a predefined frequency >> time strategies are required! Different machines can mean different data structures Different machines can mean different data structures

Aproach 1 – The Batch Report Generator - Results data is assigned to each machine that generated it data is assigned to each machine that generated it each phase of each machine is tracked (for scada system events are just digital type variables) each phase of each machine is tracked (for scada system events are just digital type variables) every alarm called during the batch is tracked and assigned every alarm called during the batch is tracked and assigned the user can see, in the report, if all the process is ok or it had any problem the user can see, in the report, if all the process is ok or it had any problem few seconds/minutes after the scada system has finished registering the production batch, the program is able to generate reports of the machines batch, connect them and print it all. few seconds/minutes after the scada system has finished registering the production batch, the program is able to generate reports of the machines batch, connect them and print it all. It can show alarms, analog variables linked to any machine, customize them and remake the report with any configuration the user wants. It can show alarms, analog variables linked to any machine, customize them and remake the report with any configuration the user wants.

Aproach 2 We decided to use this aproach in all those cases we needed a really stable application with high performance needs. We decided to use this aproach in all those cases we needed a really stable application with high performance needs. We decided to create basic modules and then add new features (other modules) when we needed We decided to create basic modules and then add new features (other modules) when we needed The 1° effort was done to wrap iFix (scada) dlls and use them with python (this was useful also for aproach 1!) The 1° effort was done to wrap iFix (scada) dlls and use them with python (this was useful also for aproach 1!) Then we started testing it for an application that comunicates with the scada framework and manage alarms/events generated by it and for an app that downlownd buffered data buffered in plc buffered and syncronises itself with it. Then we started testing it for an application that comunicates with the scada framework and manage alarms/events generated by it and for an app that downlownd buffered data buffered in plc buffered and syncronises itself with it.

Aproach 2 (2) - Events File Queue Printer Queue … User Queue SCADA Applications

Aproach 2 - Results Comparing with similar c applications performance python was not that slow… (only in some rare cases but we can do better…) Comparing with similar c applications performance python was not that slow… (only in some rare cases but we can do better…) We need to use threads… We need to use threads… Python is much, much, much more flexible! Python is much, much, much more flexible! Probably we’ll dedicate sources to improve what we obtained Probably we’ll dedicate sources to improve what we obtained

Conclusion Python worked well for us Python worked well for us Python reduced about 50% of time required for development Python reduced about 50% of time required for development UI take much of our time UI take much of our time Still some pitfalls ( apps distribution, track modules dependecies, documentation*) Still some pitfalls ( apps distribution, track modules dependecies, documentation*)

THANKS