MES in the cloud

Key Considerations for Deploying MES in the Cloud

May 31, 2018

You’ve done the math and have decided that you want to deploy your shiny new manufacturing execution system (MES) in the cloud. Doing that really depends on what you mean by “MES” and what you mean by “cloud,” as both are umbrella terms that cover a broad set of configurations. Let’s explore.

 

The cloud provides many attractive qualities for corporate IT procurement: the spend shift from CAPEX to OPEX, the lower cost of entry, better maintenance, disaster recovery and security, hardware lifecycle management, transparent hassle-free application updates and much more. (Those of you who work in regulated industries such as pharma and nuclear will understandably question the transparent application updates, but we will examine that topic another day).

 

It’s important to note that the cloud is not just a computer at a data center (or even lots of computers at a data center); rather, it’s a secure, elastic, pooled platform with immediate access and control. It’s a way of enabling infrastructure as a service (IaaS) and software as a service (SaaS), and it’s a way of enabling it better than you can.

 

Worried about security in the cloud? To make the most of elastic scaling, guaranteed uptime, metered charging and so on, you need to deploy your MES in a multi-tenant environment. Allowing your confidential IP to exist alongside that of your competitors’ might be a step too far, but by walling yourself off in a single-tenant setup, you forego many of the cloud’s advantages.

 

Then there’s your MES itself. An MES isn’t an indivisible, atomic system. It’s made up of sub-components that interact with each other and with the outside world to perform a service. Which of those sub-components can you afford to lose if your internet connection fails? Some? All? Probably not “none,” to be fair.

 

Also, how quickly do you need to react to changing operating conditions in your plant? Do you have closed-loop or advanced control? And if so, does it operate as a standalone, or is it tweaked and tuned in real time? For example, if you’re using our process sequencer software, you’re transitioning between grades under orchestrated control, reacting to process conditions in real time and working hand-in-glove with your advanced controller. Losing that functionality in the middle of a grade change would be inconvenient in the extreme.

 

The ISA-S95 view of an MES is something like this diagram on the right:

 

  • Start with a scheduled set of production orders. This frequently comes from an ERP, but ideally you’d like to build an optimized schedule, taking into account the actual situation in plant and the real behavior of your process.

  • The schedule is presented to the operators, who use their judgement to decide what to do and when.

  • The order is executed. (It’s a little more involved than this makes it sound!)

  • Process data is continually collected and “historized.” In batch or semi-batch processes, key events are detected, marking distinct phases of production.

  • Tools are provided to help you make sense of it all. In today’s connected world, you might want to bring all your data and analysis tools together across your entire enterprise. The cloud can really help with this.

 

Parts of the MES could live in the cloud and likely survive a reasonable amount of interruption. If you can’t respond to a quality incident and discover the identity of a dodgy intermediate batch right away, you’ll probably be OK (although this will change as your systems integrate closer to those of your customers). If you can’t download process orders from SAP and execute them, you’ll have a serious problem on your hands unless you have a means of creating and starting orders locally.

 

I would argue that you need some or all of the yellow components to reside on-site if you’re going to comfortably survive an appreciable outage, and/or want fine control of your production. You need to be able to create and execute orders, and you need to be able to collect process data and react quickly to process conditions.

 

In the end, there are no hard and fast rules. Is the cloud secure and stable enough? Yes. Is it up to the job of hosting a mission-critical production execution system? For the most part, yes; but it may not be ready for all of it if you want a minimum level of function in an emergency.  

 

 

To learn more about how manufacturing organizations in all industries are utilizing MES to reduce cost and increase productivity in real time, register for our upcoming webinar.

 

There was a problem storing your subscription

Leave A Comment

Comments

  • 6 years ago

    Hahaha love the title! Also the content is well mixed with data and information!

  • 6 years ago

    Hahaha love the title! Also the content is well mixed with data and information!