Do Son 
Modern industrial control systems can be a tremendous boon when it comes to productivity. But they can also present something of a weak point when it comes to cybersecurity. When these controls are compromised, an entire facility might be brought to a halt. Preventing this doomsday scenario might be considered a major priority for manufacturers. But doing so means not only thinking about conventional cybersecurity concerns, but those that stem from Operational Technology.
Why cyber resilience looks different in the OT environment
Operational Technology is a term used to refer to hardware and software that control things in the real world. So, you might think of an Industrial Control System that coordinates a production line of robots. These devices might do this via Programmable Logic Controllers and SCADA (that’s Supervisory Control and Data Acquisition) systems.
These devices aren’t as patchable as traditional IT equipment, and often they’re highly specialized and long-lasting. This presents a different kind of challenge.
Build resilience into industrial architecture and connectivity
In order for a facility to be resilient, it will need to be built from the ground up with resilience in mind. This often means segmentation. By breaking up a network, we can control the degree to which it can be compromised by a single attack. You might think of logical segmentation, which uses access control rules to keep threats contained, and physical segmentation, which creates physical boundaries.
Effective resilience isn’t just a series of updates issued to a server; it’s often built into the industrial controls themselves.
Make security part of the OT technology lifecycle
Security should be a primary concern whenever anything is being changed. Procurement criteria might include encryption, access controls, and logging. The manufacturer’s reputation for long-term support might be assessed, as might compatibility with legacy systems.
Certain updates and patches might confer a risk. These should be tested before they’re deployed, so that we can be sure they support operational objectives.
Design for disruption, isolation and recovery
You might be tempted, when assessing a facility and contemplating improvements, to wonder how an incident might be prevented. This is a sensible question. But it’s probably one best followed by another: How can operations continue when something goes wrong? By posing this question, you’ll find yourself drawn to the likeliest points of failure, and how an incident can be met effectively.
Above all, we should think not of protecting individual devices, but entire processes.