Given the significant impact these technological developments will have on the future, the future security environment will depend on how AI and analytics are integrated into a comprehensive and resilient security program that encompasses both cyber and physical security.
Cybersecurity - Industrial Internet of Things
As for how to build that overall security program, cyber surveillance technologies that can give manufacturers asset inventory and network visibility are a good place to start. As companies become more and more reliant on the digital environment, it's becoming increasingly important to have that holistic view of security. If an attack occurs within a decade similar to the major power outages that struck Ukraine twice, or the ransomware attack that struck Norwegian aluminum giant NorskHydro, companies will need to have back-up plants ready to be able to operate manually to stop an attack if necessary.
The increased connectivity of industrial systems is almost a foregone conclusion as the IoT becomes more relevant to industrial operations over the next 5 to 10 years, and as industrial systems are connected to 5G networks that can dramatically reduce the latency of communications between devices. IoT device security is often inherently weak, so when IoT devices are deployed on a large scale, industrial systems are faced with the challenge of managing device security.
Cybersecurity-industrial operations
Worse still, increased connectivity means more hackers can try to break into systems, higher-end hackers may be able to snoop on systems, and cybersecurity issues worsen as connectivity grows. Moreover, many industrial systems could injure people if manipulated in a particular way, so increased connectivity affects not only the management and protection of industrial systems, but also public ***** policy making.
Cybersecurity-Digital Transformation
The biggest impact suffered by industrial cybersecurity will be the unintended consequences of digital transformation. Digital transformation is great and necessary, but it comes with risks. As we introduce more and more digital endpoints, data streams ensue. The data streams will grow by leaps and bounds beyond our ability to process and effectively analyze all of the data on-site. Moreover, we will use this data to drive decisions about the process, or even to drive the process itself. Eventually, we may begin to feed these analytical data products back into the process through AI/machine learning.
In other words, the process generates the data, the data leaves the process network and flows out into the clouds, fog, lakes, field, outside, etc., to be analyzed, reused, and fed back into the process. All of this introduces new risks to the process data and related systems external to that control/process network in ways that we've only just begun to consider.