By now, you know you want to create an automation roadmap that aligns with corporate goals, which typically center on production, efficiency or quality gains. In my last couple of blogs (accessible here and here), I provided a structured approach to developing your automation roadmap, which generates a nice list of projects, in order of priority, accompanied by budget and timeline considerations. But before you make your way to the corporate offices to garner budget approval, there are a couple of additional considerations that will add the finishing touches to your automation roadmap: network architecture and safety. In this post I’ll address network architecture; next month I will address safety.
Often overlooked in the manufacturing environment is the importance of a sound network architecture. When I go into manufacturing facilities, it is pretty common to see networks that were seemingly pieced together with parts from the local electronics store, CAT5 cables routed through the rafters, and switches sitting out in the open collecting dust. My intent here is not to knock these systems. When they were put in, the technology was sound, the demands on the system were often low, and the systems were probably not performing any critical functions. It was cheap, served its purpose, and was miles ahead of other network standards that were available.
Architecture that is pieced together, however, is simply not going to cut it in the future. The demand for data is going to be too great because network traffic is only going to grow and the information that the data provides will become critical to operations. Network architecture must be taken seriously and designed to reliably handle a much larger bandwidth.
It is important to understand that I am not an advocate for collecting every data point, every millisecond, and dumping it all into a historian database. I recognize there are certain heavily regulated industries that have very exhaustive data collection and archiving requirements. Many industries, though, that do not have such requirements. Either way, having access to every bit of data is extremely important.
Outside of regulated industries, I would never suggest starting out building a historian database, but rather a system that monitors the data and derives real-time, actionable information. Regardless of what you do with the data, it is transmitted on the network and eats up bandwidth.
Consider what your data needs will be down the road and ensure each of your projects supports a network architecture that can handle that data. You can build the network as you go, where each project adds an additional node. In this fashion, you avoid a complete overhaul of your network, which could prove difficult to sell from an ROI standpoint, especially if your company does not have an appreciation for what data, as actionable information, can do for the operation.
A good first step is to conduct a network assessment to gauge the current status of your network. Next, you need to determine your network’s future needs, which include who needs access, where, and what restrictions need to be in place. This will likely require you to engage with your IT department to understand what you can and cannot do. It is better to have this conversation early in the process, rather than after you have started installing network gear. If you do not have information on your future data needs, or are unable to get it, a safe bet is to plan on connecting your entire line via Gigabit network with several segments that allow you to tie to other networks. This will give you a solid network backbone to build on.
There are many important aspects of integrating a sound network architecture. I would encourage you to engage an automation solution provider that has this expertise. A qualified company will know the right questions to ask and be able to design a system that will suit your needs now and well into the future. Once this is done, it is easy, and likely not overly expensive (depending on your needs) to integrate with your automation roadmap. This kind of forethought will impress those making the decisions on capital investments.
Michael Gurney is CEO of Concept Systems Inc., a Certified member of the Control System Integrators Association. See Concept Systems’ profile on the Industrial Automation Exchange.
