If you’re like most producers, your answer is likely “yes.” Producers in the process industries have a keen interest in analytics for the problems they can help address.
Most analytics solutions that try to solve these types of production problems took data from a variety of sources that had been difficult to connect. The Connected Enterprise vision can help enable better business insights because of its broad approach.
Similarly, supporting technologies are broadening their footprint. Instead of integrated solutions running on a level three workstation (outside of the control layer, but within the plant boundary), current and future platforms will include advanced technologies in control devices, cabinet appliances or in a secure cloud platform.
This leaves many producers asking themselves, “Where should my analytics solution run?”
Consider the following as you explore the options of the three industrial layers - the device (control LAN) layer, system (workstation) layer and cloud (internet-enabled layer, which is frequently, but not always off-premise).
Where is the Data?
For most producers, answering this question is a matter of knowing what information can help solve the problem and where is it housed today.
Imagine the control system pyramid, with some indication of boundaries where data exchange is controlled for security purposes.
Where is data aggregated in the format that any specific analytics solution will need?
Where and How Can You Provide Acceptable Industrial Security?
If an application is going to turn equipment on or off, producers want to know that a hacker hasn’t changed the decision or presented wrong data that drives to decisions.
Secure data exchanges between isolated data layers within production environment systems exist for this reason.
Higher levels on the control pyramid are more open and lower levels are more secure and isolated. However, if data goes only in one direction, then answers will only be available far from devices.
Short-term opportunities can be lost due to lags and waiting on manual responses, and in the end, decisions may need to be based on data only available in that perceived “less secure” layer.
Today’s software can provide higher levels of security that can be expanded as the capabilities of technology and threats increase over time.
While this isn't an answer to the core question, you may want to consider driving toward the highest security at all layers, remaining open to data exchange and applying available secure technologies to help protect users with multi-layered security.
Where there are security protections at each layer of The Connected Enterprise, producers should take into account site security standards and implementation levels when designing analytics solutions.
How Does the Location of the Analytics Solution Affect Cost?
Pricing will differ based on the layer of execution, differences in costs of secure data exchange, and authentication requirements needed.
Delivery engineering, product and support costs will also vary based on where an application is running. For example, with a cloud platform, a vendor may charge for resources on a recurring cost structure.
However, this means local IT won’t have additional computers or devices to maintain, update or manage. On the other hand, device management software has made great strides, making delivery of hardware updates easier.
Therefore, considering the cost of the options that will help you achieve your targeted results is another aspect that will help your decision.
Ultimately, there is no right answer to where your analytics solutions should run.
But for many producers in the process industries, answering these three questions will help you arrive at a viable solution.