Is Data Centricity the New Frontier in Cybersecurity?
Lane Long posted on January 30, 2019 |
As IIoT ecosystemsdevelop, companies concerned about security could begin to look inward to protect their datasets. (Image courtesy of Strategic Finance.)
As IIoT ecosystems develop, companies concerned about security could begin to look inward to protect their datasets. (Image courtesy of Strategic Finance.)

Among all the developing technologies with the potential to revolutionize manufacturing today, perhaps none is regarded with more excitement than the Industrial Internet of Things (IIoT). From labor allocation to supply chain management, increased connectivity promises to help optimize manufacturers across nearly every aspect of their operations. With great opportunity, however, may also come considerable risk. Higher levels of interdependence and growing numbers of linked devices are making networks more vulnerable even as they develop and become more valuable to manufacturers. As the IIoT comes of age, cybersecurity will be increasingly difficult—and important—for these companies to manage.

Real-Time Innovations (RTI) recently hosted a webinar on the risk of cyber attack to businesses that are becoming increasingly reliant on the IIoT. A leading IIoT connectivity framework software provider, RTI held the discussion to summarize the scope of the challenge. Taking the discussion a step further, CEO Stan Schneider also offered listeners a tantalizing piece of the solution. According to Schneider, comprehensive security frameworks that gird both edge-of-network and data flow entry points will be essential for stopping catastrophic network breaches in the future.

Internet of Targets?

Just as an airport relies on multiple layers of security to ensure safety, IIoT security architects should be prepared to integrate defenses at every level. (Image courtesy of RTI.)
Just as an airport relies on multiple layers of security to ensure safety, IIoT security architects should be prepared to integrate defenses at every level. (Image courtesy of RTI.)

Schneider began the webinar with the observation that cybersecurity in the era of the IIoT is in some ways analogous to the physical security at airports. There are many connected layers of security that must hold up individually as well as collectively. A single miscue can compromise the integrity of an entire network, which can take a factory offline in a matter of seconds in the event of an attack. These stakes are increasing as the IIoT grows more potent. Schneider emphasized that virtually all “critical infrastructure” will be on the grid in the future, which makes the cost of any security risk difficult to quantify.

The sheer number of connected devices in the factories of the future will make companies vulnerable. Not only are the devices themselves dramatically different, but they also likely use a wide array of platforms, operating systems and software. This gives prospective cybercriminals an abundance of targets. A single inconsistency can be exploited to compromise an entire system, which means that companies can’t rely on the security of just one layer of connectivity.

Data Centricity

One concept that could be integral to solving the problem of how to secure large numbers of connected but incongruous devices is known as data centricity. This approach rests on IIoT functionality coming not through devices being linked with one another, but through the devices being linked with the data they need to optimally function. The concept is perhaps best visualized by considering a large IIoT ecosystem, characterized by disparate devices on disparate operating systems. The devices are not connected with one another, but they pull data from the exact same dataset. The company-wide dataset that all tools draw from in order to perform is what connects the devices.

Data centricity hinges on the availability of the shared dataset being perfect to all applications at all times. (Image courtesy of RTI.)
Data centricity hinges on the availability of the shared dataset being perfect to all applications at all times. (Image courtesy of RTI.)

In spite of its streamlining nature, this universal data access poses a significant security risk when taken alone. Making all data available to all devices at all times is like a call for attention from cyberattackers. The crucial counterweight, to Schneider and RTI, is a focus on data security rather than delivery security.

Dataflow Control

The technique espoused in this webinar doesn’t involve any coding. There are no APIs—it’s designed solely to control the means by which devices interact with the flow of data, rather than focusing on the security of the network edge or the conduits through which data flows. These superficial layers, which Schneider denotes as “Level 1” or “Level 2” security, are much more complex and much less vital to protect than is the actual exchange between devices and data. From an architectural standpoint, it should also be noted that this control mechanism performs independently of specific operating systems.

Data defense security also differs from Level 1, 2 or 3 security in that it’s less vulnerable to a complete system breakdown even if an attack were to occur. If a dataflow “pipe” were to become compromised, for instance, the actor could theoretically penetrate other IIoT-connected nodes adjacent to the point of entry in any direction. If the data itself is compromised, however, the progression of an attack will be somewhat linear and predictable. The upshot is that there’s often a clear terminal point. If a node that is reached by “bad data” has no other outgoing connections, the bad data stops there.

Securing Dataflow to High-Leverage Applications

RTI’s webinar carefully notes the importance of a layered defense architecture, a concept it refers to as “defense in depth.” There is no substitute for a comprehensive system in which edge applications (including physical hardware), hosting and networks are all adequately secured. But a separate component—data—is most important and impactful of all as a cybersecurity focus area.

This is especially true in industries where momentary failure can have outsized consequences. If real-time performance is essential, a disruption of data flow—or even a simple corruption of data—can be disastrous. While data centricity may indeed make defending any industry’s IIoT landscape more streamlined, such security may have an even simpler value proposition than the one Schneider lays out. For companies in sectors like aerospace, autonomous vehicles and medicine, the consequences of failure could be too difficult to overcome.


Recommended For You