ES
Secure Remote Access for Industrial Control Systems

Secure Remote Access for Industrial Control Systems

Guide to implementing secure remote access for SCADA and PLC systems covering VPN architectures, jump servers, session recording, and IEC 62443 compliance.

Published on December 30, 2025

Secure Remote Access for Industrial Control Systems

This guide explains how to implement secure remote access for SCADA and PLC systems, covering VPN architectures, jump servers, session recording, micro-segmentation, and IEC 62443 compliance. It synthesizes industry standards, vendor approaches, and practical deployment steps so automation engineers can design, validate, and operate remote access that preserves availability while minimizing cyber risk.

Key Concepts

Understanding the fundamentals reduces risk and aligns remote access with operational requirements. The following concepts form the foundation for secure remote access in industrial control systems (ICS).

Zero Trust and Outbound-Only Connectivity

Zero Trust means never implicitly trusting network location or device identity. For ICS remote access, this translates to outbound-only VPN or TLS tunnels from the OT site to a broker or gateway, avoiding inbound firewall ports on control networks. Many standards and vendor solutions now prescribe outbound TLS/IPsec tunnels with strong cryptography (2048-bit RSA key exchange and SHA-256 or better) to prevent unsolicited inbound connections and simplify firewall rules (see design guidance in industry best practices).

Jump Servers and Controlled Hops

Jump servers (also called bastion hosts or jump boxes) provide an audited, hardened intermediary between remote operators and control equipment. Best practice requires:

  • Mutual authentication (client and server certificates),
  • At least one controlled hop from the jump server into the OT network,
  • Use of unprivileged operator accounts and unique session credentials, and
  • Automatic secure disconnection or physical session termination after the session completes (as recommended by ACSC guidance) [2].

Micro-Segmentation and Network Zoning

Micro-segmentation reduces attack surface by enforcing per-asset or per-protocol policies instead of broad trust across an OT subnet. IEC 62443 defines zones and conduits; remote access must traverse defined conduits with controlled services, firewall rules, and deep packet inspection where appropriate [1][4].

Authentication, Authorization and Accounting (AAA)

Implement multi-factor authentication (MFA) with a hierarchy that can include PKI (X.509), time-based one-time passwords (TOTP), hardware tokens, or certificate-based device authentication. Role-based access controls (RBAC) should enforce least privilege and map to IEC 62443 SR 2.x requirements for account management [1].

Session Recording and Forensics

Recordings should capture timestamps, source identity, session commands, configuration changes and screen output. Integrate session recording with SIEM or audit repositories. Virtual desktop infrastructure (VDI) combined with session video capture provides both operator visibility and secure evidence for post-incident forensics [1][6].

Standards and Compliance

Remote access designs must satisfy regulatory and standards-based requirements. Key references include IEC 62443, NIST SP 800-82r1, ACSC protocols, and vendor/industry guidance.

  • IEC 62443: Requires zoned architectures, role-based access control, encrypted communications, and audit logging for control system remote access (see SR 1.x–4.x and conduit/zone mapping) [1][4].
  • NIST SP 800-82r1: Recommends segmentation, no direct vendor inbound access to control equipment, use of jump hosts, and MFA for SCADA and PLC remote connections [9].
  • Australian Cyber Security Centre (ACSC): Provides a procedural remote access protocol with explicit steps including time-limited access, MFA, mutual authentication, and ensuring vendor accounts are unprivileged and temporary [2].
Standard Key Remote Access Requirements
IEC 62443 Zoned access, encrypted tunnels, role-based controls, session auditing [1][4]
NIST SP 800-82r1 Segmentation, no inbound ports, MFA, jump hosts [9]
ACSC Protocol Outbound VPN, mutual authentication, physical disconnect, unprivileged accounts [2]

Implementation Guide

Successful implementation follows a phased, auditable approach: assess, design, pilot, deploy, and validate. Below is a practical step‑by‑step guide with technical specifics.

1. Initial Assessment

Inventory assets and communications. Identify PLCs, RTUs, HMI versions, remote diagnostic ports and legacy protocols (Modbus, DNP3, Profinet). Capture telemetry such as CPU/IO load and deterministic timing constraints. Determine which assets can tolerate additional network latency (target under 150 ms for control loops where possible) and which must use local control only.

2. Risk Analysis and Zoning

Map the control network into IEC 62443 zones: safety, control, and enterprise/DMZ. Define conduits for remote access and enforce them with next‑generation firewalls and micro-segmentation appliances. Define explicit firewall rules per asset and protocol—avoid any rule that broadly permits “all traffic” from a remote provider.

3. Choose Architecture and Tools

Select vendor solutions that meet IEC 62443 principles and operational constraints. Options include edge brokers (ei3, Claroty, Phoenix Contact), hardware gateways (Waterfall), and managed Zero Trust Remote Access platforms. Ensure chosen solutions support:

  • TLS 1.2/1.3 or IPsec with 2048-bit RSA and SHA-256 minimum,
  • Mutual certificate authentication (PKI/X.509),
  • Session recording and export to SIEM, and
  • Integration with existing identity providers and MFA systems.

4. Pilot Deployment

Pilot on a small set (5–10 machines) to validate connectivity, latency, logging, and failover behavior. Configure strict access windows, MFA, and ephemeral credentials for the pilot. Run planned maintenance tasks and emergency scenarios to validate operator workflows.

5. Full Rollout and Hardening

Scale the deployment, apply host and network hardening, and disable default accounts. Implement device posture checks to prevent unmanaged devices from connecting. Apply patch management procedures consistent with operational availability constraints.

6. Validation and Ongoing Testing

Validate policies continuously. Use periodic penetration testing, red team exercises, and automated configuration checks. Monitor for anomalies and ensure session recordings are retained per retention policy (commonly 1–3 years depending on regulatory requirements).

Architecture Patterns and Recommendations

This section highlights proven architecture patterns and vendor approaches you can adopt based on the asset mix and operational requirements.

Zero Trust Remote Access (ZTRA)

ZTRA brokers all access, enforces per-session authorization, and applies micro-segmentation down to the asset level. Many modern OT vendors provide ZTRA implementations that integrate with legacy PLCs using edge collectors to translate proprietary protocols while preserving policy enforcement and auditing [3][6].

Outbound VPN / Broker Model

OT edge devices initiate outbound VPN or TLS sessions to a cloud or on-premises broker—this removes the need for inbound firewall ports at the OT site and supports remote connectivity over Wi‑Fi, 4G LTE, or wired links. Configure TLS with current cipher suites and disable older weak algorithms; prefer TLS 1.3 when supported [5].

Air-Gapped and Unidirectional Gateways

Where absolute isolation is required, unidirectional gateways (data diodes) permit data flow out of the control network while blocking inbound traffic entirely. For remote maintenance, pair data diodes with a dedicated, monitored jump server in a DMZ and strict human-in-the-loop authorization [7].

Solution Type Strengths Considerations
Zero Trust Broker (ei3, Claroty) Asset-level policies, micro-segmentation, session recording Requires edge agents/collectors; integration effort
VPN Gateway (Phoenix Contact) Simple outbound tunnels, works with legacy PLCs Ensure MFA, logging, and least-privilege policies
Unidirectional Gateway (Waterfall) Strong isolation; prevents inbound attacks May limit remote control capability; increased complexity

Session Recording, Auditing and Forensics

Auditability provides both deterrence and incident response capability. Effective session recording captures:

  • Session start/end timestamps and duration,
  • Authenticated user identity and authentication method (MFA, certificate),
  • Source IP address and device posture status,
  • Commands executed, configuration changes, and file transfers, and
  • Screen video or virtual desktop snapshots when available for investigator context [1][6].

Integrate recordings into a central SIEM with retention policies and immutable storage to satisfy audit and legal requirements. Ensure that session capture does not unduly increase latency or interfere with deterministic control loops.

Best Practices

Adopt these operational best practices to balance cybersecurity with operational performance.

  • MFA and PKI: Enforce MFA for all remote connections; use device and user certificates (X.509) for mutual authentication. Where possible, implement just-in-time (JIT) keys and short-lived credentials [3].
  • No Direct Vendor Access: Do not allow vendors inbound access to control devices. Require proxy/jump host access with operator presence where mandated by NIST and ACSC guidance [9][2].
  • Least Privilege: Implement RBAC aligned to job function and IEC 62443 SR 2.x controls. Use unprivileged operator accounts for remote sessions and require elevation only for specific, audited tasks [1].
  • Time-Limited Access: Grant access for set windows and enforce automated disconnects. Audit and require re-authorization for extended sessions [2].
  • Segmentation and Conduits: Enforce zone-to-zone conduits per IEC 62443 and keep IT/OT traffic path‑separated with strict firewall rules and DPI where required [1][4].
  • Device Posture and Allowlisting: Verify endpoint posture before allowing access and employ allowlists for management tools and protocols [8].
  • Phased Rollout and Training: Pilot the solution, then expand; train both OT engineers and remote vendors in the access procedures and incident workflows [3].

Product Solutions and Compatibility

Several vendors offer OT-focused remote access solutions. As of 2025–2026, leading approaches emphasize OT-native brokers, compatibility with legacy PLCs, micro-segmentation, and session recording:

  • ei3 Secure Remote Service: ZTRA with micro-segmentation and just-in-time keys; integrates with legacy assets via edge connectors [3].
  • Phoenix Contact Remote Access: VPN-based, single-port tunnelling, aims to minimize cloud dependency and ease integration with industrial networks [4].
  • Claroty SRA: Granular asset policies, protocol awareness, automatic disconnects and deep auditing for SCADA/PLC systems [6].
  • Waterfall Security Gateway: Encrypted low-latency tunnels and unidirectional options for high-assurance isolation [7].

When selecting products, verify support for the specific PLC/HMI protocols in your plant and confirm cryptographic configurations, certificate lifecycle management, and logging export capabilities.

Deployment Checklist

Use this checklist to ensure consistent deployments that meet operational and regulatory requirements.

  • Complete asset and protocol inventory; document timing sensitivities.
  • Define IEC 62443 zones and conduits; map remote access paths.
  • Select technology that supports outbound-only connections and mutual authentication.
  • Implement MFA and certificate-based device authentication.
  • Configure jump servers with session recording and RBAC.
  • Apply micro-segmentation rules and per-asset firewall policies.
  • Conduct pilot on 5–10 devices; run functional and failure-mode tests.
  • Integrate logs and recordings into SIEM; set retention and alerting rules.
  • Train OT personnel and vendors; publish access procedures and emergency contacts.
  • Schedule periodic revalidation, patching windows, and penetration tests.

Incident Response and Operational Considerations

Prepare a focused incident response plan for remote access compromises. Key elements include:

  • Real‑time alerting on anomalous access patterns from SIEM,
  • Immediate revocation mechanisms for certificates or credentials,
  • Forensic preservation of session recordings and network captures, and
  • Reconstitution procedures for safe recovery (e.g., isolate affected segments, failover to manual/local control) [8].

Operational teams should also plan for network outages: ensure remote access solutions support failover communication channels (e.g., LTE fallback) and provide clear escalation and on-site contingency procedures.

Summary

Secure remote access for industrial control systems requires a combination of architecture, policy, and operational discipline. Implement outbound-only encrypted tunnels, jump servers with mutual authentication, session recording, and micro-segmentation. Align deployments with IEC 62443 and NIST SP 800-82r1 guidance, adopt MFA and PKI, and pilot changes before full rollout. Together, these measures reduce attack surface while preserving the operational access necessary for maintenance and emergency support.

For hands-on assistance with design, pilot implementation, and compliance validation, contact our engineering team to discuss requirements, architecture reviews, and managed deployment services.

References and Further Reading

Frequently Asked Questions

Need Engineering Support?

Our team is ready to help with your automation and engineering challenges.

sales@patrion.net