Smart manufacturing and smart factories are the embodiment of digital transformation where industrial automation and data-driven processes are embraced to meet the demands of modern businesses and supply chains. Cybersecurity, meanwhile, is often a tradeoff that’s made during any such fundamental transformation.
As we’ve seen with the attacks against refrigeration controllers across Europe, however, any shortcoming around operational technology (OT) cybersecurity can have devastating economic and societal impacts. Connectivity has turned once-isolated manufacturing processes into an accessible threat vector for determined hackers with geopolitical, social, or for-profit motivations.
Smart manufacturing operations are connecting industrial control systems (ICS), robotics, Internet-of-things (IoT) sensors and, and other production equipment to cloud services or the public internet for third-party access. The business and competitive benefits are real for factories, including the introduction of artificial intelligence (AI) and large-language models (LLMs) to rapidly analyze manufacturing data and process efficiencies.
Cybersecurity leaders responsible for industrial automation and smart factory cybersecurity understand that a compromised fleet of programmable logic controllers (PLCs), or unavailable human-machine interfaces (HMIs), can quickly interrupt production, damage equipment, compromise product quality, and create unsafe working conditions. These disruptions not only impact a factory’s bottom line, but may also disrupt downstream supply chains. Industrial automation cybersecurity must therefore protect both digital assets and the physical operations they support.
This blog will examine:
How digital transformation expands smart factory attack surfaces
The security risks behind modern manufacturing
Provide a checklist of industrial automation security priorities CISOs
Industrial automation cybersecurity must evolve quickly while not slowing digital transformation or undermining its productivity gains. It’s a tricky balance for security teams.
CISOs are not able to reject the connectivity demands of factory automation. OT environments are increasingly sending data not only to the enterprise network but also outside the network edge to cloud-based services for processing and analysis. Third-party integrators, vendors, and partners require access to these environments for analytics, predictive maintenance, and overall service management, creating new exposures and external risks that must be managed.
We saw the attack pathways that exposed refrigerator controllers provided in Europe; default passwords and poor visibility and configuration management may have played a part in the costly compromises of these assets. These attacks should serve as a wake-up call for other industries, including manufacturing about the blurring of traditional IT and OT boundaries. Security teams have to worry about legacy, unsupported, decades-old OT assets, but also whether they’re now safely connected to the internet, or the security posture of third parties with access to HMIs and SCADA systems at the heart of industrial automation.
The challenge for CISOs is that digital transformation often progresses faster than the security programs designed to protect it. New technologies are introduced, legacy systems remain in production, and connectivity expands without always being accompanied by appropriate security controls.
Connectivity in manufacturing environments has brought to the surface two industrial automation cybersecurity problems that are putting smart factories at risk.
The technical debt paradigm around OT is well understood: Decades-old technology and unsecured protocols persist inside manufacturing and other critical sectors that cannot easily be updated, patched, or replaced without threatening the integrity of industrial processes or physical safety. PLCs, field devices such as sensors and actuators, and terminal units are at the forefront of technical debt issues in factories. Because firmware and software vulnerabilities cannot easily be patched, for example, expensive compensating controls must instead keep attackers at bay.
Despite the efforts of CISA and pressure from security experts, movements such as secure by design and secure by default take time and commitment from original equipment manufacturers (OEMs) to implement these changes, and from integrators to deploy them in a timely manner.
With AI uncovering vulnerabilities at an unprecedented pace, security teams are running into much shorter update cycles and longer windows of exposure, further exposing the technical debt issue many factories are facing.
Access is step one in any cyberattack, and often it’s enough to allow an attacker to manipulate settings on an internet-exposed asset, disrupt processes, or damage equipment. In a smart factory, exposed PLCs, HMIs, engineering workstations and other automation components can provide attackers with opportunities to conduct reconnaissance and compromise those and other manufacturing systems.
Many of these assets rely on legacy industrial protocols that were designed for trusted environments rather than today’s threat landscape. Protocols such as Modbus, DNP3 and other industrial communications technologies were not necessarily developed with modern authentication, encryption or access-control capabilities in mind. When exposed or improperly secured, they can provide opportunities for unauthorized access, manipulation of process communications or disruption of operations.
These risks are compounded by incomplete asset inventories. Manufacturers lacking visibility into connected assets can not adequately deploy compensating controls such as network segmentation, threat detection, or secure access technology.
Industrial automation systems must be resilient in the face of cyberattacks. CISOs must assume a mindset that breach and exploitation are inevitable, and architect the protection of business-critical assets for operational resilience. The strategy includes five key aspects:
Asset inventories of devices central to industrial automation must be accurate and continuously updated. This includes technology starting at the lowest levels of the Purdue Model for ICS and providing decision-makers with information such as model and serial numbers for PLCs, HMIs, engineering workstations, and IoT devices. Inventories should also include firmware versions, operating systems, communication protocols, vulnerabilities and network relationships. All of this information informs the reminder of the smart factory cybersecurity program.
The ability to segment OT assets into zones that are governed by distinct security policies, if so desired, is essential to protecting industrial automation. Factory assets have many dependencies, and these represent potential attack pathways that can also be abused by an attacker with illicit access to the OT network. Segmentation can be used to isolate specific assets, reducing the possibility of lateral movement between assets and other parts of the network, and as a result limit the blast radius of a compromise. Zero-trust principles can further strengthen these boundaries by requiring explicit authorization for communications and access.
Manufacturing environments operate on dozens of distinct, proprietary OT and IoT protocols that require OT-aware network monitoring and threat detection. These tools are better suited to identify anomalous protocol communication, unauthorized changes, or activity that strays from established baselines. Operational context is key to a healthy detection strategy, which should inform security teams in order to distinguish genuine threats from legitimate engineering activity.
A recent Claroty survey of 2,000 security and technology leaders asked respondents about the tangible risks to operations posed by unmanaged and unsecured third-party access. The responses were eye-opening, including the fact that three-quarters said their organizations had suffered at least one incident impacting operations related to third-party access, and half said they had only partial or no monitoring of third-party connections.
Legitimate third-party access belonging to integrators or partners are potential paths for attackers onto manufacturing networks. Organizations should eliminate unnecessary persistent access, enforce multifactor authentication, use controlled OT-specific secure access solutions, restrict permissions and monitor remote sessions.
The days of vulnerability management based on CVSS criticality scores are gone. Manufacturing organizations must determine the business impact of critical automation assets and use that information to prioritize remediation of those devices according to exploitability, active attacks in the wild, and operational impact. Compensating controls such as segmentation, access restrictions, enhanced monitoring and secure configuration can reduce overall exposures.
Smart factories will continue to evolve, bringing greater automation, connectivity and operational intelligence to manufacturing. Cybersecurity must evolve alongside them. The objective is not to slow digital transformation or undermine the productivity gains it promises. It is to ensure that increased connectivity does not introduce unacceptable operational risk.
For CISOs, this requires closer collaboration with manufacturing engineering, operations and business leadership. Security decisions must account for production priorities, equipment limitations, safety requirements and business continuity.
Interested in learning about Claroty's Cybersecurity Solutions?
Life, uninterrupted
We maximize your availability, strengthen your insurability, and support compliance to ensure operational resilience.
We’re here to help you secure your OT environment.
We’re here to help you secure your OT environment.